Flint Mining in Prehistoric Europe Inter-1 PDF

European Association of Archaeologist, 12th annual meeting
Cracow, Poland, 19th-24th September 2006
Flint mining
in the Prehistoric Europe
interpreting the archaeological records
Edited by
Pierre Allard, Franoise Bostyn, Franois Giligny and Jacek Lech
BAR International Series 1891

2008
BAR
European Association of Archaeologist -12th annual meeting Cracow, Poland, 19th-24th September 2006
Flint mining in the Prehistoric Europe: interpreting the archaeological records

the individual authors 2006
ISBN
Editing: Dominique BOSSUT

(Institut National de Recherches ArchologiquesPrventives, Direction Nord-Picardie)
Secratary of edition: Franoise Bostyn

(Institut National de Recherches ArchologiquesPrventives, Direction Nord-Picardie)
The texts have been re-written in part for enhanced clarity and smoothness by Christine Merli-Young
Printed in England by
12TH Annual EAA meeting, Cracow 2006 - Flint mining in Prehistoric Europe: interpreting the archaeological records
Table of Contents
Flint extraction and processing from secondary flint deposits in the north-east of Scotland
in the Neolithic period...................................................................................................................................... 1
Alan Saville
Flint working at the early linearbandkeramik settlement of Geleen-Janskamperveld........................... 13

Marjorie E. Th. de Grooth
An economy of surplus production in the early Neolithic of Hesbaye (Belgium): Bandkeramik blade
debitage at Verlaine Petit Paradis............................................................................................................... 31
Pierre Allard, Laurence Burnez-Lanotte
The prehistoric flint mining complex at Spiennes (Belgium) on the occasion of its discovery
140 years ago.......................................................................................................................................................41
Hlne Collet, Anne Hauzeur, Jacek Lech
A new flint mine at Flins-sur-Seine/ Aubergenville (Yvelines, France)...................................................... 79

Franoise Bostyn, Franois Giligny, Adrienne Lo Carmine
The Krzemionki flint mines latest underground research 2001-2004....................................................... 97

Jerzy Bbel
Open-cast flint mining, long blade production and long distance exchange:
an example from Bulgaria.............................................................................................................................111
Laurence Manolakakis
Flint mining in early Neolithic Iberia: a preliminary report on Casa Montero (Madrid, Spain)....... 123
Marta Capote, Nuria Castaeda, Susana Consuegra,Cristina Criado, Pedro Daz-del-Ro
Intensive extraction of non-metallic minerals during the early protohistory in the northern
half of europe................................................................................................................................................. 139
Yoann Gauvry
Ideology and influences behind theNeolithic flint mines of the Southern Britain.................................. 155
Paul Wheeler
Alan Saville: Flint extraction and processing from secondary flint deposits in the north-east of Scotland in the Neolithic period
Flint extraction and processing from

secondary flint deposits in the north-east
of Scotland in the Neolithic period
Alan Saville
Abstract: Secondary flint deposits were of considerable importance to prehistoric people in areas where no such mate-
rial existed in a primary context. In locations where secondary flint deposits were available close to the surface, multiple
shallow pits were often dug over extensive areas. Intensive initial processing of the extracted flint usually took place on
site adjacent to the extraction pits, without evidence for tool manufacture. Despite the intensive nature of such extraction,
the use and distribution of the resultant flint may be only local and regional. This contribution will consider some of the
specific circumstances of such secondary flint exploitation at one particular location, Den of Boddam in Aberdeenshire,
north-east Scotland.
Keywords: Den of Boddam, flint, Neolithic, quarrying, Scotland, secondary deposits.
INTRODUCTION
In parts of Europe where there is no flint (or relevant al- available in small-size pieces, of indifferent quality, and
ternatives) available as a fresh raw material in its primary occurred sporadically and somewhat unpredictably, so its
geological context, prehistoric people inevitably sought exploitation is likely to have been seen as a high-invest-
out other locations where flint or analogous materials were ment / low-return type of activity, perhaps only undertaken
available. Probably the most common sources of such raw expediently in the course of other pursuits. It is, of course,
material were those locations where flint was exposed at an activity which is difficult to evaluate from the archaeo-
or near the surface by natural processes of erosion, such logical record. The collection of surface flint pebbles will
as on beaches, at cliffs, in river beds, and so on. This was be very unlikely to leave any direct traces at the collec-
certainly true in Scotland, where there is no flint in a pri- tion site, and the collection locations may now be invisible
mary geological context. If there ever were flint-bearing because the occurrences of flint have been exhausted or
Cretaceous chalk deposits on land they have long since because of the obscuring effects of erosion, colluviation,
eroded away. Flint is locally available, however, in derived peat growth, and other topographic changes.
and redeposited form, on beaches around some parts of
the coast and elsewhere in various glacial, glacio-fluvial On the other hand, there are in various parts of Europe
and riverine deposits (Marshall 2000; Wickham-Jones and geological situations where flint is available in abundant
Collins 1978). Flint obtained in this way was used for arte- and predictable quantities where it is not in a primary con-
facts throughout the Mesolithic and Neolithic periods and text, but has accumulated as a result of specific processes
into the Bronze Age. of erosion affecting the deposits in which it originally
formed. These so-called secondary flint deposits (Weisger-
However, the flint from such deposits was normally only ber 1987: 131), where they were accessible to prehistoric
1
Fig. 1: The flint-rich Buchan Ridge Gravel occurs in a small area south of Peterhead, in Aberdeenshire, north-east Scotland. The two known quarry sites
shown, Den of Boddam and Skelmuir Hill (Saville 1995), are approximately at the east and west edges respectively of the occurrence of the Buchan
Ridge Gravel deposits.
extractive technology, would have formed reliable sources

of raw material albeit not of top quality for prehistoric
people prepared to invest in the labour required to exploit
them.
1. DEPOSITS AT DEN OF BODDAM
One such location where a secondary flint source was

available and was exploited occurs at Den of Boddam in
Aberdeenshire, north-east Scotland (Fig. 1). Here a geo-
logical occurrence, known as the Buchan Ridge Gravel
Formation, is the result of deposits formed millions of
years ago in the Tertiary era, which have become buried
inland as a result of subsequent processes of erosion and
deposition (Bridgland 2000; Bridgland et al. 1997; Merritt
et al. 2003). As it survives today at Den of Boddam, the
variant of these deposits referred to here as the Buchan
Ridge Gravel (and abbreviated to BRG) consists largely
Fig. 2: Den of Boddam. Geological test-pit section, showing the relation- of cobbles and pebbles of flint and quartzite (and similar
ship of the Buchan Ridge Gravel to overlying deposits and details of the
solid and kaolinized clasts within the Buchan Ridge Gravel. Q = quartz- very robust rocks), thought to be the remains of a fossil
itic cobble; other cobbles and pebbles shown with a solid outline are of
marine beach, and where the BRG is overlain by a metre or
flint; those with a dashed outline are decomposed (or ghost) clasts of
other rock types. so of later glacial deposits (Fig. 2). Most other, less resil-
2
Fig. 3: Den of Boddam location map, showing the

position (and limited extent) of the main archaeo-
logical excavations in 199293. The black dots
show the position of all the extraction pits visible
on the surface, as plotted by surveyors of the Royal
Commission on the Ancient and Historical Monu-
ments of Scotland (RCAHMS 1994: 14-15). The
stream running through the Den of Boddam was
dammed in recent times to create a reservoir to
feed water to a mill further downstream.
ient, lithic components of the BRG have decomposed as a flint clasts sized 50mm or larger (taken as the arbitrary
result of deep weathering (Hall 1986) and form the matrix base-line above which size the flint pebbles and cobbles
for the surviving flint and other clasts, giving what was in are considered to have been most suitable for prehistoric
origin an open-framework deposit the character of a ma- knapping) constitute by weight approximately 35%, or
trix-supported one. Beyond this site the BRG survives as about a quarter of the total BRG deposit. To give that some
a very localized phenomenon, forming the non-continuous context, obtaining 500 flint clasts sized 50mm or larger
capping to the higher parts of undulating low-relief land, would require 75.23 kg of the BRG. However, the aver-
extending west from the modern coast over an area about age maximum dimension of the flint cobbles sized 50mm
13 kilometres (eight miles) across. or larger is only 64mm, and, on the basis of the samples
The existence of the flint-rich deposits at Den of Boddam examined, amongst 500 clasts one would expect only 12
was identified by prehistoric people, probably following cobbles to be 100mm or larger in maximum dimension.
on from their observation of a dense presence of flint cob-
bles in the stream running through this location, which is
a relict glacial meltwater channel (in which the stream is
now dammed to create a small reservoir). Prehistoric quar-
ry pits are still plainly visible on the surface as pronounced
hollows at Den of Boddam on the steep unploughed slopes
of the channel (Fig. 3), making this a unique survival of a
Neolithic industrial monument in Scotland.
Samples taken from the same level of the BRG at Den of

Boddam as was dug into in prehistory have shown that ap-
proximately 68% (by weight) of the deposit is comprised Fig. 4: Den of Boddam. Cobbles and pebbles (washed) extracted from
of solid clasts of all kinds and sizes (Fig. 4). Of these, samples of the Buchan Ridge Gravel.
3
Fig. 5: Den of Boddam. Plan of the pits and other features exposed in 199293. Pits 19, 30, 46 and 101 were completely excavated; pits 14 and 67 were
half-sectioned. Pits 19 and 30 were shallow features of unknown purpose which did not penetrate as far as the Buchan Ridge Gravel. The buried soil was
preserved intermittently between and around pits where modern cultivation had failed to reach the base of the remains of upcast quarry spoil.
Thus if prehistoric people were only interested in large high granite content) following deep weathering, and this
cobbles, extracting BRG would be rather unrewarding la- process has acted to effectively bleach the exterior and
bour. All the signs are, however, that smaller pebbles were interior of the flint cobbles. Flint is also present within the
a viable resource for the Neolithic flint knappers, and that metre-thick glacial gravel which overlies the BRG, and
the BRG was a desirable deposit to exploit for flint. There comprises the same type of rounded, chatter-marked cob-
is anecedotal information from the observation of deep bles, but these are distinctively different in appearance,
sections into the BRG at Den of Boddam and elsewhere in having a dark brown to black cortex and a grey-brown to
the local area that the availability of larger-sized clasts in- brown interior. In those areas investigated by excavation
creases with depth, but the Neolithic miners were unable, at Den of Boddam it is clear that for the most part this flint
given the quarrying technology being employed, to access from the glacial gravel was ignored, since debitage which
the deposits very deeply. is brown in colour represents only a tiny proportion of the
total recovered.
All the pebbles and cobbles have a rounded form and
chatter-marked cortex surfaces (Fig. 4). Internally the flint Neither in size nor in quality, therefore, is the Den of Bod-
is only rarely clear-structured and free of flaws. Coarse dam flint suitable for the manufacture of axeheads or other
inclusions, unsilicified areas, cavities and faults of various large implements, only for making smaller flake imple-
kinds are common. Flint from the BRG at Den of Bod- ments such as scrapers, knives and arrowheads. The rela-
dam is basically light grey in colour, both internally and tively few flint axeheads found in north-east Scotland are
externally, reflecting the grey-white colour of the BRG as virtually all made of flint imported from outside the region
a whole, which is kaolinized (on account of its original- (Saville 1994, 1999).
4
Fig. 6: Den of Boddam. Pit 46 north-south section drawing, showing the position of the birch charcoal sample which produced a radiocarbon age of
3100-2900 cal BC (OxA-13103; table 1). The sample was located at the base of a substantial infill deposit of flint-knapping debris. The buried soil inclu-
sions within the fill on the south side of the pit indicate an area where the pit edge has collapsed over an undercut into the Buchan Ridge Gravel.
2. QUARRYING clear signs of edge collapse (Fig. 6). This is unsurprising

given the instability of the deposits, especially as there are
Although the existence of the flint-bearing BRG deposits indications in many cases at Den of Boddam that when
at Den of Boddam must have been recognized since the first dug the pits were bell-shaped (that is they expanded
Later Mesolithic period, from when there exist the first at the base, presumably to maximize the yield of flint per
signs of human inhabitation of the local area, it is not until pit and because it was the flint from the BRG which was
a developed stage of the Neolithic that invasive exploi- required, and not that from the overlying glacial gravel).
tation by quarrying takes place. The extraction method There is an inevitability, when dealing with pits in unstable
employed involved the repeated excavation of roughly deposits, that the excavated shapes of the pits will gener-
circular, cylindrical pits through the overlying topsoil ally bear only a partial relationship to their original form
and glacial gravel and down into the BRG (Fig. 5). The (Budziszewski 1997).
deepest pits (amongst the very small sample of pits inves-
tigated thus far) reached just over 4m below the surface, The prehistoric exploitation of the BRG was thus a mat-
and even this must have been high-risk for those extract- ter of simple extractive technology, but extraction which
ing the flints, because the BRG and glacial gravel deposits would have required careful management of spoil if ex-
are inherently unstable. At Den of Boddam the absolute traction capacity was to be controlled and in any sense
thickness of the BRG is unproven, but it is known to con- maximized. In some cases the fact that there are gaps be-
tinue with an equal density of flint for many metres below tween what appear to have been perfectly productive pits
the depth of the deepest prehistoric pits. In virtually every may indicate the position of former spoil heaps; in other
instance the pits examined archaeologically have shown cases the pits overlap or were contiguous (Fig. 5).
5
The scale on which extraction took place at Den of Bod- cobbles found among the backfill deposits in the pits, but
dam was considerable. There is still the surface evidence even if one were to assume that only half of the number of
for at least 458 pits (Fig. 3). Following archaeological in- cobbles extracted were subject to, at the minimum, some
vestigations (Bridgland and Saville 2000; Saville 1995), testing, and that only half the number of those went on
it is estimated that perhaps as many as 1000 pits were to be knapped to some extent, this would still be in the
originally dug at this location over an area of almost 12 order of 3000+ cobbles sized 50mm or over per quarry pit.
hectares (30 acres). In crude spatial terms this would make Compared to the serendipity of finding flint cobbles from
the extraction at Den of Boddam more extensive than at local beaches and chance exposures, the resource available
any of the English flint mines on the Chalk, except Easton from quarrying BRG must have seemed well worth while.
Down, Wiltshire (Barber et al. 1999: 58), though this is
not comparing like with like. Grimes Graves in Norfolk Dating the quarrying activity at Den of Boddam has proved
may only have somewhat over 400 shafts across an area of difficult, both in relative terms in the absence of diagnostic
eight hectares, but of course the shafts there relate to inten- lithic tool-types or other material culture, such as pottery,
sive underground mining and the amount of high-quality and in absolute terms because of the lack of organic ma-
usable flint extracted at Grimes Graves must exceed that terials for radiocarbon dating (the acidic deposits are not
at Den of Boddam by an enormous factor. (Note that the conducive to the preservation of items of bone, antler, or
extent of the mined area at Grimes Graves is disputed; for wood). Charcoal is present on the site but only in a very
example, Lech and Longworth [2000: 70] claim it to be 25 few instances during excavation were samples of charcoal
hectares.) recovered from useful secure contexts where the amount
or size of the charcoal precluded the possibility of con-
Very approximate calculations can be made to estimate the fusion resulting from post-depositional movement. How-
yield from the Den of Boddam quarry pits, on the basis of ever, two secure samples, both of birch charcoal, taken
the very small sample of the site which has been archaeo- from infill horizons within quarry pits, give two termini
logically investigated. Ignoring the superficial soil horizon ante quos of c. 3000 cal BC for the quarrying activity in
and the glacial gravel which overlies the BRG, the area of one part of the site (see Table 1). The position of one of
BRG extracted from a quarry pit would be in the region, these samples, at the base of an infill horizon of knapping
rather conservatively estimated, of six cubic metres. On the debris, is probably at the surface at what was a relatively
data provided by the BRG samples discussed above, this swift initial collapse and infill of the lower part of the
amount of BRG would be likely to contain some 12,720 pit (Fig. 6), implying that the date for the extraction pit
flint cobbles sized 50mm or larger. Unfortunately the ex- might not be very much earlier than the radiocarbon date.
cavations at Den of Boddam have provided no data on Some confirmation for this might be shown by radiocar-
the number or percentage of the numerous unstruck flint bon dates from an area of buried soil at Den of Boddam
Lab code Site coding and Sample type Context 13C 14

C years BP Cal date BC OxCal
and no. sample no. v3.10 @ 95.4%
GU-3438 DB91/9002 buried soil / humic fraction Area 1, ditch section -29.3 4580 60 35203090
GU-3439 DB91/9002 buried soil / humin fraction Area 1, ditch section -28.9 4530 50 33703030
OxA-13102 DB91/357 birch charcoal ditch Area A west; pit in- -26.9 4372 35 30902900
fill within knapping debris
OxA-13103 DB93/323 birch charcoal Area 3, Pit 46, base of con- -25.1 4387 34 31002900
text 48 (knapping debris)
Table 1: Radiocarbon dates from Den of Boddam.
6
sealed beneath dumps of upcast material from the quarry Experience of digging through the BRG while excavating
pits, which gave a result in the second half of the 4th mil- showed that, particularly when conditions were wet, the
lennium cal BC (Table 1). matrix became very sticky and clinging, adhering to the
clasts to the extent that it was often difficult to discriminate
There are no chronological indicators for the length of between flint and non-flint, let alone between which flint
time over which quarrying took place at Den of Boddam, cobble might be more amenable to hand knapping than an-
so although from the number of pits we have an idea of the other. It might of course be the case that in such conditions
large scale on which quarrying took place, we have no idea (which would be dangerous) the Neolithic quarriers might
of the intensity of this activity. If the real total of pits at this either not work at all, or they might stockpile the clasts to
location is 1000, then 5 pits dug per year would represent allow them to weather-off naturally. The fact that quartz-
a 200-year span, 10 pits dug per year a 100-year span, 20 itic cobbles occasionally show signs of knapping, and that
pits a 50-year span, and so on (and this assumes that pits flint cobbles were occasionally used as anvils, suggests
were dug every year, which is of course another unknow- that confusion did sometimes exist.
able factor). Equally, there is no archaeological evidence
which can help indicate the number of people who might Soft-hammers of wood or antler would not survive in the
have been involved in the processes of quarrying and soil conditions at Den of Boddam, but in fact the debitage
knapping at Den of Boddam, and for how much time per left on site shows virtually no indication of anything other
year they devoted to these tasks. So at the moment there than hard-hammer flaking. The hammers are predominant-
is no factual basis for assessing which year span might be ly cobbles of quartzite and analogous hard stones and only
appropriate, but it might be guessed that 20 pits per year rarely of flint. They vary in size and the extent to which
(yielding 250,000+ cobbles) would be beyond the raw ma- they have been used, some being abandoned while still re-
terial requirement, if not beyond the manpower resources taining the hammer surface(s) intact, others after clearly
deployable. having spalled and fractured during use (Fig. 7).
3. PROCESSING Some hammerstones have been used as anvilstones (or vice

versa). Anvilstones, almost always non-flint cobbles, also
In the area investigated archaeologically at Den of Boddam vary widely in size and extent of usage. The pits which
the flint extracted from the quarry pits was processed by result from knapping can be single, minimal, and hardly
primary knapping immediately adjacent to the pits. Cob- visible, perhaps indicative of use for opening only one
bles and pebbles were tested and either immediately dis- flint cobble, or multiple and deep from repeated usage, and
carded if considered inappropriate for additional working, these stones frequently broke while in use (Fig. 8). The an-
or they were subjected to further knapping, occasionally vilstones are common over 240 were recovered from the
to the extent of producing exhausted cores. Some cobbles recent excavations at Den of Boddam but at this site their
were opened by free-hand knapping using a hammerstone, use appears primarily to have been in the testing and open-
with the cobble being held in the hand while struck, oth- ing of cobbles, rather more than for bipolar core knapping
ers were opened by the anvil technique. In the latter case as such, since the evidence for bipolar reduction amongst
the flint cobble was placed and held on a cobble anvilstone the debitage is limited. Conventional platform cores (Fig.
(usually a quartzitic cobble) before being hit with a stone 9), and sub-discoidal, sub-Levallois cores are the two most
hammer. The deciding factors involved in the choice of common methods of flake production. The later type often
opening technique are unknown; logically it might be ex- approximates to the tortoise core form, whereby a corti-
pected that the anvil technique would be employed for the cal, often primary, flake from the edge of a cobble (whether
more rounded cobbles and the freehand knapping for the anvil struck or not), is trimmed around part of the periphery
more elongated or sub-angular cobbles which already prof- by removals struck from the ventral, bulbar surface, then a
fered a usable platform or on which one could easily be flake (or sometimes more than one) is removed across the
created. On the other hand there could have been personal ventral surface, having a faceted platform determined by
preference for one technique over another on the part of the previous peripheral flaking (cf. Saville 2006).
individual knappers.
7
Fig. 7: Den of Boddam. Hammerstones. Fig. 8: Den of Boddam. Anvilstones.
Archaeological investigations at Den of Boddam have re-

sulted in the recovery of hundreds of thousands of pieces
of flint debitage, but have not yet discovered any finished
(or even roughout versions) of implements which might be
seen as the intended product of this large-scale enterprise.
There are occasional ad hoc tool-forms such as minimally
retouched scrapers and irregular pieces with areas of sec-
ondary retouch, but no recurrent types of formal imple-
ments. Probably the most parsimonious explanation for the
absence of finished implements or roughouts / preforms at
this quarry site is that the intended product on-site was the
flake blank, and that further processing of the blanks took
place off-site, or at least in a different part of the site from
any which has yet been investigated. This would explain
both the absence of the implements themselves, and also
the absence of successfully-produced, reasonably well-
Fig. 9: Den of Boddam. Platform core. sized and well-proportioned flakes amongst the residual
debitage. The most likely end product from those flakes
is considered to be arrowheads (Saville 2005), but for the
moment this must remain a suggestion rather than a fact.
8
Specific fieldwork and the study of museum collections Bridgland, D.R., Saville, A., and Sinclair, J.M., 1997. New
to track the Neolithic distribution of Buchan Ridge Gravel evidence for the origin of the Buchan Ridge Gravel, Aber-
products has not taken place in any rigorous fashion, but deenshire. Scottish Journal of Geology 33, 43-50.
casual observation would suggest that this flint was not
dispersed far beyond the north-east of Scotland. The distri- Budziszewski, J., 1997. Mine 1/4 of the Za Garncar-
bution would appear to be an easterly one, from down to- zami mining field in Ozarow (Central Poland): remarks
wards Dundee in the south to around Inverness to the west on methods of studying shallow flint mines. In A. Ram-
(i.e. a distance of perhaps 130 km [80 miles] in each di- os-Millan and M.A. Bustillo (eds.), Siliceous Rocks and
rection beyond the in situ Buchan Ridge Gravel deposits). Culture, Granada: Universidad de Granada, 151-162.
It is not possible, however, to discriminate between flint
which may have been obtained directly from extraction Hall, A.M., 1986. Deep weathering patterns in north-
pits at Den of Boddam or elsewhere, and flint which has east Scotland and their geomorphological significance.
been derived from the Buchan Ridge Gravel and obtained Zeitschrift fr Geomorphologie 30, 407-422.
from natural exposures or superficial deposits, around the
coast, in river beds, or elsewhere. This unsatisfactory situ- Lech, J. and Longworth, I., 2000. Kopalnia krzemienia
ation is of course compounded by current ignorance of the Grimes Graves w swietle nowych badan. Przeglad Arche-
actual intended outcome of the extraction activity in terms ologiczny 48, 19-73.
of implement types. It is hoped that ongoing studies will
be able to throw more light on the matter. Marshall, G., 2000. The distribution of beach pebble flint in
western Scotland with reference to raw material use during
the Mesolithic. In S. Mithen (ed), Hunter-gatherer Landscape
Alan Saville Archaeology: the Southern Hebrides Mesolithic Project 1988-
Archaeology Department, 98, Vol.1, Cambridge: McDonald Institute, 75-77.
National Museums Scotland
Chambers Street, Edinburgh EH1 1JF, Scotland, UK Merritt, J.W., Auton, C.A., Connell, E.R., Hall, A.M. and
e-mail: a.saville@nms.ac.uk Peacock, J.D., 2003. Cainozoic Geology and Landscape
Evolution of North-East Scotland. Edinburgh: British Ge-
ological Survey.
References
RCAHMS, 1994. Monuments on Record: Annual Review
Barber, M., Field, D. and Topping, P., 1999. The Neolithic 1993-4. Edinburgh: Royal Commission on the Ancient and
Flint Mines of England. Swindon: English Heritage. Historical Monuments of Scotland.
Bridgland, D.R., 2000. Discussion: the characteristics, Saville, A., 1994. Exploitation of lithic resources for stone
variation and likely origin of the Buchan Ridge Gravel. tools in earlier prehistoric Scotland. In N. Ashton and A.
In J.W. Merritt, E.R. Connell and D.R. Bridgland (eds.), David (eds.), Stories in Stone, London: Lithic Studies So-
The Quaternary of the Banffshire Coast and Buchan: Field ciety, Occasional Paper 4, 57-70.
Guide, London: Quaternary Research Association, 139-
143. Saville, A., 1995. Prehistoric exploitation of flint from
the Buchan Ridge Gravels, Grampian Region, north-east
Bridgland, D.R. and Saville, A. 2000. Den of Boddam. Scotland. Archaeologia Polona 33, 353-368.
In J.W. Merritt, E.R. Connell and D.R. Bridgland (eds.),
The Quaternary of the Banffshire Coast & Buchan: Field Saville, A., 1999. An exceptional polished flint axe-head
Guide, London: Quaternary Research Association, 102- from Bolshan Hill, near Montrose, Angus. Tayside and
115. Fife Archaeological Journal 5, 1-6.
9
Saville, A., 2005. Prehistoric quarrying of a secondary flint

source: evidence from north-east Scotland. In P. Topping
and M. Lynott (eds.), The Cultural Landscape of Prehis-
toric Mines, Oxford: Oxbow Books, 1-13.
Saville, A., 2006. Flint technology and production asso-

ciated with extraction sites in north-east Scotland. In G.
Krlin and G. Weisgerber (eds.), Stone Age Mining Age,
Bochum: Deutsches Bergbau-Museum (= Der Anschnitt
19), 449-454.
Weisgerber, G., 1987. The technological relationship be-

tween flint mining and early copper mining. In G. de G.
Sieveking and M.H. Newcomer (eds.), The Human Uses
of Flint, Cambridge: Cambridge University Press, 131-
135.
Wickham-Jones, C.R. and Collins, G.H., 1978. The sourc-

es of flint and chert in northern Britain. Proceedings of the
Society of Antiquaries of Scotland 109 (1977-78), 7-21.
Acknowledgements
I am grateful to Franoise Bostyn for the initial invitation

to speak in the flint mining in Prehistoric Europe session
at the EAA Conference in Krakow in 2006, and for her en-
couragement and patience while awaiting the written ver-
sion for this volume. Figs 1 and 3 were drawn by Marion
ONeil, Figs 2, 5 and 7 by Alan Braby, and Fig.6 and the
photographs are by the author. I wish to thank the Royal
Commission on the Ancient and Historical Monuments of
Scotland for permission to incorporate the results of their
survey work in Fig.3, and Craig Angus for help with im-
ages.
10
Marjorie E. Th. de Grooth: Flint working at the Early Linearbandkeramik settlement of Geleen-Janskamperveld
11
12
Flint working at the Early Linearbandkeramik

settlement of Geleen-Janskamperveld
(province of Limburg, the Netherlands)
Marjorie E. Th. de Grooth
Abstract: The main objective of this study was to investigate the way flint working was organized at the Early LBK set-
tlement of Geleen-Janskamperveld, thus getting an insight into the procurement strategies and the technological choices
made at the beginning of Bandkeramik settlement west of the river Rhine. A new evaluation of raw material character-
istics made it possible to determine the (probable) extraction site, located around 22 km south of the settlement. In the
Banholtergrub, a dry valley 5 km to the east of the Rijckholt flint mines, nodules originating from the upper Cretaceous
Lanaye chalk layer were extracted from residual loams. Intensive processing took place in the settlement; surplus material
was distributed in the shape of partially reduced blade cores and blades. A comparison with the contemporary settlements
at Elsloo (Graetheide plateau) and Langweiler 8 (Aldenhovener Platte) made it clear that different exchange mechanisms
prevailed, so that Langweiler 8 may have received both prepared cores from JKV and blades from Elsloo.
Keywords: Linear Bandkeramik, flint, procurement strategies, distribution networks, Graetheide, Rhineland.
Introduction
The Bandkeramik settlement Geleen-Janskamperveld subsequent analysis, combining data on decorated ceram-
(JKV), situated on the well-known Graetheide Plateau ics and on the structure of houses, arrived at a subdivision
in the Netherlands (Fig. 1), was excavated in 1990/1991. into four housing generations (Van de Velde et al. 2007).
An initial analysis of house plans, decorated ceramics and After a marked hiatus, the site was resettled during Mod-
settlement structure is given by Louwe Kooijmans et al. dermans phase IIc, but the extent and spatial organisation
(2003), and De Grooth (2003b) reported on the flint as- of this phase of occupation cannot be assessed.
semblage. The full excavation report will be published The main objective of the study presented here was to
shortly (Van de Velde et al. 2007). investigate the way(s) flint working was organized at
The settlement extended over approximately 4.5 ha, of the Early JKV site, thus getting an insight into the pro-
which 2.7 ha, or 61% was investigated. Habitation started curement strategies and the technological choices made
early in the Flomborn phase of the Linearbandkeramik at the beginning of Bandkeramik settlement west of the
(LBK). This first occupation lasted approximately 100 river Rhine. For this purpose, a total of 7941 flint arte-
years, comprising the phases Ib and Ic of the Dutch chro- facts, weighing c. 58 kg, were analysed (De Grooth 2007).
nology (cf. Modderman 1970). During this time, an esti- 4864 of these were recovered from 42 pits containing at
mated 90 houses were constructed, of which 69 have been least 15 flints, and dated to the Flomborn phase by their
excavated. Originally, on the basis of the decorated ce- ceramic content (Table 1). These figures differ from those
ramics, Van de Velde distinguished five phases within this published previously (De Grooth 2003b), partly because
Early LBK habitation (Louwe Kooijmans et al. 2003). A material from some of the excavation trenches were not
13
Fig. 1: The location of Geleen-Janskamperveld (JKV) and other sites under discussion. Map based on Modderman 1970, Fig. 1.
Type JKVE1 JKVE2 JKVE3 JKVE4 All Early

N % N % N % N % N %
Cores/hammerst 13 1.3 11 0.8 18 1.0 5 0.7 47 1.0
Hammerstone fragm 27 2.8 28 2.1 43 2.4 11 1.6 109 2.2
Crested blades 10 1.0 7 0.5 11 0.6 4 0.6 32 0.7
Rejuvenation flakes 22 2.3 29 2.1 46 2.5 9 1.3 106 2.2
Flakes with cortex 213 21.8 261 19.2 356 19.6 138 19.6 968 19.9
Flakes without cortex 354 36.2 466 34.2 693 38.1 219 31.1 1732 35.6
Blades with cortex 34 3.5 41 3.0 46 2.5 19 2.7 140 2.9
Blades without cortex 114 11.7 125 9.2 166 9.1 91 12.9 496 10.2
Tools 102 10.4 129 9.5 182 10.0 96 13.6 509 10.5
Chips (<15 mm) 55 5.6 244 17.9 226 12.4 98 13.9 623 12.8
Artificial blocks 33 3.4 22 1.6 32 1.8 15 2.1 102 2.1
Total 977 100.0 1363 100.0 1819 100.0 705 100.0 4864 100.0
Table. 1: Geleen-Janskamperveld: amounts and percentages of artefacts through time, for Early LBK habitation (ceramically dated pits
containing at least 15 flint artefacts).
14
initially available for study and partly because some of the and medium-sized (between 1 and 10 mm) black or very
pits were assigned to a different phase after re-analysis of dark grey round, ovoid or sickle-shaped spots; dark wisps
the decorated ceramics. or tendrils.
-Small angular cavities.
1. Raw material procurement
For the most part, flints from the other levels within the West-
The flints used at JKV originated almost exclusively in the ern Lanaye Member have the same characteristics, notably
western facies of the Lanaye Member of the upper Creta- as regards colour, texture and the presence of both dark and
ceous Gulpen Formation (Felder and Felder 1998; Felder light spots and specks. The detailed study of four of them
et al. 1998, Felder and Bosch 2000), occurring in the lime- showed the following divergences: dense concentrations of
stone area south of the river Geul, at a distance of 20-30 small white spots are present in layer 01; the flints from lay-
km from the settlement. This region is considered to be the er 2A often contain seemingly overlapping inclusions, and
major procurement area for Bandkeramik settlements in fossils that are not completely silicified; abrupt transitions
the Graetheide cluster (Bakels 1978, de Grooth 1987), as in colour occur in the laminated areas in layer 12A. Then,
well as for the Rhineland (e.g. Zimmermann 1995, Zim- the cortex may be thick (e.g. layer 02A, 12A) or irregular,
mermann et al. 2004). The Lanaye stratum comprises 23 with hollows and protuberances (layer 01).
different flint-bearing levels (numbered from bottom to The combined presence of several kinds of dark inclusions
top). Of these, layer 10, comprising the largest amount of is the major characteristic distinguishing Lanaye flints
usable flint, is the one most extensively exploited during from other flint types in the region (De Grooth in prep.).
the Neolithic. Lanaye flints (commonly referred to by ar-
chaeologists as Rijckholt flint), display a large variability. Lanaye flint nodules could have been extracted from four
The nodules in general have a length, width and thickness different depositional contexts:
of at least 30 cm. The cortex is thin, rough and whitish. -Firstly, from their primary position, in the chalk bedrock.
Natural fracture planes are often covered with iron incrus- -Secondly, from slope, talus or scree deposits, that came
tations. The internal colour varies from very dark to very into being when valleys developed during the Pleistocene
light grey, both sometimes with a hint of blue. The surface cut into the chalk beds, thus exposing and eroding them.
of the artificial fractures is smooth, but dull, the texture -Thirdly, from residual loams (also known as eluvial de-
is mainly fine-grained. Both colour and texture may vary posits) that are the result of disintegration of the chalks
within individual nodules; the lighter grey parts often con- during the Tertiary.
tain zonated areas, with gradual transitions. Sometimes -Finally, from gravels deposited by the river Meuse dur-
the zone directly under the cortex is the darkest, with a ing the Pleistocene and the Holocene.
smoother, more vitreous texture. The main types of inclu-
sions are (Felder et al. 1998, De Grooth 1998): The conditions prevailing in these secondary deposits
-Concentrations of light (white or light grey) round sometimes led to visible alterations in the aspect of the
specks (<1mm); isolated small (13) and medium-sized flints. Flints from slope deposits differ from the material
(310 mm) round or ovoid spots, light grey or white, with in primary context only in regard to a slight weathering of
the same texture as the matrix. the cortex and other natural surfaces. Material collected
-Small (13 mm), medium-sized (3 10 mm) and large in river gravels may be identified as such only when parts
(>10 mm) ringed spots, round or irregular in shape, with of the cortex or other natural surfaces are still present, as
a smooth whitish outer ring and a smooth or rough darker the river transport leads to heavy abrasion of cortex, to a
centre. decrease in size of the flints, and to an increase in the fre-
-Large (>10 mm) spots, round or irregular, abrupt bor- quency of non-cortical natural surfaces. These often have
der, light grey or whitish, with a texture rougher than the a battered aspect and carry a glossy patina (pebble patina,
matrix. Verhart 2000). The most extensive alterations, however,
-Large, vaguely lighter grey flecks. are present on flints embedded in residual loams, especial-
-Concentrations of black round specks (< 1mm); small ly when these are mixed with iron-rich Oligocene sands.
15
They were found to differ from primary Lanaye material by the excavations in the Schone Grub, a dry valley in the
in the following aspects (De Warrimont and Groenendijk north-western part of the prehistoric mining area (Rade-
1993; De Grooth in prep.): makers 1998; Felder et al. 1998).
-Differences in the colour and texture of the cortex. -2. The steep slopes between the Upper Terrace and the
-Changes in colour, as a result from the infiltration of iron Middle Terrace surrounding the plateau known as De
compounds. This often manifests itself as a reddish brown Kaap, located some 500 m further to the south, may also
zone directly under the cortex and/or as yellowish streaks have been exploited (Felder 1998). At present these activi-
that penetrate deeply into the nodules. In extreme cases, ties cannot be dated, but some Bandkeramik adzes have
the whole nodule is affected. been found on the Rijckholt plateau (Brounen and Peeters
-The presence of a thick (>1 mm) white zone directly un- 2000/2001).
der the cortex. Early Neolithic exploitation of residual flint deposits may
-Natural and artificial fracture planes with a dusty as- have taken place at the following sites (cf. Felder 1998):
pect, due to the presence of dense concentrations of minus- -3. Hoogbos, between Mheer (mun. Margraten, NL) and
cule, vermiculate, spots. s-Gravenvoeren (mun. Voeren, B). Exploitation is thought
-Concentrations of white specks (< 1mm) and small white to have taken place in the steep valley slope, where the ma-
spots (indicating the presence of flints from Lanaye layer terial crops out. Nothing is known about the age and char-
01 and/or the underlying Lixhe Member). acter of the mining activities, but Louis (1936) reported
-The degree of translucency. Using a method devised by that some of the cores and rejuvenation tablets collected
Ahler (1983), unweathered freshly knapped flakes were here resemble those found at Omalien (i.e. Bandkeramik)
studied in a darkened room with a light source (provided by sites in the vicinity of Lige. In colour, texture and inclu-
a 12 V/20W halogen desk lamp) diagonally behind them. sions this material is identical to the flints encountered at
The boundary between the opaque and translucent parts Rijckholt; the cortex, however, is rough, thin and brown-
of the pieces was then marked in pencil, and the thick- ish. Yellowish streaks are infrequently present, as are con-
ness measured with a pair of sliding callipers. Later on, centrations of small light spots. Opaque reddish brown
the measurements were grouped into 5 classes: T1: trans- zones are very rare. The translucency is low to medium.
lucency 2.4 mm (opaque); T2: translucency between 2.5 -4. Banholtergrub, close to Banholt (mun. Margraten,
and 4.9 mm (low); T3: translucency between 5.0 and 7.4 NL), situated on the northern slope of a narrow dry valley.
mm (medium); T4: translucency between 7.5 and 9.9 mm The eluvial deposits lie on the remaining chalks and are
(high); T5: translucency greater than 10.0 mm (very high). covered by Quaternary gravels. Some Oligocene sands are
Whereas the material from the Rijckholt mine shafts and mixed with the loams (W.M. Felder, oral communication
the slope deposits both displayed a low to medium trans- 29/06/2006). Recently, Brounen and Peeters (2000/2001)
lucency, those residual sites containing a mixture of loams presented plausible evidence for (open-cast) mining and
and Oligocene sands produced highly translucent flints. knapping activities at this site during Early Neolithic,
more specifically Bandkeramik times. Again, the material
Prehistoric extraction points, that probably were in use is similar to the Rijckholt sample. The cortex, however,
during the Early Neolithic are known for three of the four is rough, thin, brown or grey. Frequently a thin reddish
depositional contexts (Fig. 2, Table 2). brown, glass-like zone below the cortex is present, and a
-1. Lanaye nodules from both a primary context and from thick white layer may occur; brown or yellowish streaks
slope deposits may have been extracted in the mining area are common, as are concentrations of light specks (<1
situated between Rijckholt (mun. Eijsden, NL) and Sint- mm); the fracture surface lacks the dusty appearance typi-
Geertruid (mun. Margraten, NL). The well-known under- cal for Rullen (see below). The translucency is medium to
ground mining activities there are contemporary with the high.
Michelsberg culture (Felder et al. 1998). Earlier mining -5. A cluster of extraction sites at Rullen and Sint
of primary material, by means of shallow pits, cannot be Pietersvoeren (Vrouwenbos according to Felder 1998;
excluded a priori. In the immediate vicinity, flints were also known as Bois Communal and Bois des Sapins) in
extracted from slope deposits as well, as has been shown the municipality of Voeren (B). The material extracted
16
Primary Slope Residual Oligocene Oligocene

context deposits loams sands cobbles
Rijckholt mines X
Rijck. slopes/Kaap X
Hoogbos X
Banholt X X
Rullen c.s. X X X
Rodebos X X X
Table. 2: Geological characteristics of known extraction points.
Fig. 2: Possible Early Neolithic extraction sites. Map: Zuid-Limburg, 1:50.000 (topografische dienst Nederland 1980).
17
here is commonly known as Rullen flint. At these sites, intense, this results in the honey or egg-yolk coloured
the residual loams are mixed with important amounts of nodules often seen as the typical Rullen flint (esp. Lhr et
Oligocene sands, rich in iron oxides, displaying intense al. 1977). In many artefacts found at the Rullen and Sint
red and yellow colours. In 1998, during a rescue exca- Pietersvoeren sites, however, solid reddish- or yellowish-
vation prior to the construction of a liquid gas pipeline, brown colours are limited to the outer part of the nodules,
traces of a funnel-shaped extraction pit, with a preserved whilst the rest of the piece is grey, with just some brown,
depth of ca. 300 cm and a reconstructed diameter of 7,20 yellow or orange streaks.
m, were found at Rullen (Vermeersch et al. 2005). The -6. Rodebos, close to Remersdaal (mun. Voeren, B). The
available radiocarbon dates correspond to the end of the site is undated and the character of the mining activities
Neolithic or even later. The presence of blanks and tools unknown. The geological situation is identical to that at
in earlier Neolithic settlements, however, points to ex- Rullen; i.e. residual loams mixed with Oligocene sands.
tensive extraction activities in this period: Rullen flints Besides the basic Lanaye attributes, this material is charac-
are encountered sporadically during the LBK, but are an terized by the frequent presence of a reddish brown glassy
important to predominant raw material in settlements of zone under the cortex. Sometimes a thick white zone oc-
the Grossgartach, Planig-Friedberg and Rssen cultures curs as well. Specks, yellowish wisps and fracture planes
(Gehlen and Schn in press). At the Rullen/Sint Pieters- with a dusty aspect are rare. The fracture surfaces are more
voeren sites, materials originally deriving from all the reflective than those at Rijckholt and Rullen, and the trans-
levels within the Lanaye Member were exploited. This lucency is extremely high: almost 60% of the sample be-
follows from the observation that concentrations of light longs to class 4 and 5.
specks (<1 mm), the size of a pins head, and concentra-
tions of small light spots are common, whilst abrupt tran- No specific extraction points for gravel flints are known,
sitions in colour and sharply defined stripes parallel to but the material was widely used during the Neolithic, es-
the cortex occur as well. The cortex is brown, sometimes pecially in areas where it formed the nearest source of raw
rough, sometimes smooth; mostly thin, but sometimes material (Weiner 1997; Mischka 2004).
thick. A thick, white layer is often present, especially in
material from the Sint Pietersvoeren (Bois Communal By comparing the raw material characteristics of the dif-
and Bois des Sapins) sites. The fracture planes are dull ferent extraction points with those of the JKV flints, the
and mostly have a dusty aspect, probably because the latters probable origin could be established.
bleaching has made the presence of dense concentrations Firstly, their slightly weathered cortex shows that the La-
of minuscule, vermiculate, spots visible. As befits their naye flints worked by the inhabitants of Geleen did not
origin in the Western Lanaye deposits, the basic colour of originate from a primary geological position. Nor were
the flints found at the different Rullen/ Sint Pietersvoeren they collected in river gravels. The majority seem to de-
extraction sites was grey. Three phenomena, however, rive from a secondary position, either from slope deposits,
caused alterations, that make it possible to distinguish such as found in the vicinity of the Rijckholt mines, or
these flints from the Lanaye material from primary chalk from residual loams.
deposits. Firstly, their long stay in an eluvial matrix gave Secondly, in a sample of 358 well-dated JKV artefacts, 82
them a bleached aspect, possibly caused by the dissolu- pieces (23 %) showed a vitreous reddish-brown zone un-
tion of the carbonates that are present in large amounts derneath the cortex; 50 times (14 %) a thick white zone
in primary Lanaye material (cf. MacDonnell et al. 1997; was present under the cortex, and 16 artefacts (4.5%) had
Giot et al. 1986 describe the same loss of carbonates for yellowish-brown streaks. The characteristic features of the
flints from Le Grand-Pressigny). This bleaching some- Rullen flints (solid yellow or brown colours, concentra-
times affected the dark inclusions as well. Then, possibly tions of small light round spots and dusty fracture surfaces)
because of the same loss of carbonate content, Rullen were absent in the sample. Thus, the JKV material shows
flints have a high translucency (predominantly class 4-5). characteristics not encountered in the Rijckholt or Rullen
Thirdly, infiltration of the iron compounds present in the samples, but resembles best the raw material collected at
matrix led to a yellowish-brown discoloration. At its most Banholtergrub and Rodebos (Table 3).
18
White zone Red zone Brown streaks Dusty surface

JKV 14 % 23 % 4.5 % No
Rijckholt No No No No
Hoogbos No Rare Rare No
Banholt Common Common Common No
Rullen c.s. Common Frequent Frequent Frequent
Rodebos Common Frequent Rare Rare
Table. 3: Comparison of raw material characteristics between JKV and extraction points.
%
80,0
70,0
60,0
50,0 T1
T2
40,0 T3
T4
30,0
T5/6
20,0
10,0
0,0
)
4)
94
7)
)
8)
5)
94
5)
0
)
4
9
=2
=5
=1
=1
=1
=1
=1
=2
=1
(N
N
(N
N
(N
N
(N
N
(N
t(
t(
(
s(
n
p
V
ol
os
+1
ol
,5
lle
aa
bo
JK
kh
nh
+2
gb
V
eK
Ru
de
jc
JK
Ba
oo
V
Ro
Ri
D
JK
Fig. 3: Comparison of translucency between JKV and extraction points.
19
Finally, the translucency of 271 JKV artefacts was meas- (De Grooth 1987), Lige-Place St. Lambert (Cahen 1984),
ured. 3% were opaque, 46% slightly translucent, 35 % or Verlaine-Petit Paradis (Allard 2005). The toolkit used
showed a medium translucency and 16% were highly by JKVs inhabitants is extremely conventional, both in
translucent. Moreover, 181 of the artefacts (67%) were composition and in morphology (Table 4). Arrowheads,
completely translucent (Fig. 3). This phenomenon partly borers (or rather: inserts for drilling machines), truncated
reflects the relatively low thickness of the JKV artefacts, blades, blades with lateral retouches, end-scrapers and
but it also points to an overall high translucency of the raw side-scrapers are the main standardized tool types. Follow-
material used. In order to make a comparison with the data ing common practice, to these are added the blades with
from the extraction sites possible, the measured translu- intensive gloss, interpreted as sickle inserts, even though
cency was presented in three ways: 1. raw data; 2. increase they often are not modified by intentional retouch. Stand-
with one mm for all entirely translucent artefacts; 3. in- ardized tools were almost exclusively made from blades
crease with 2.5 mm. This, of course, leads to an increase of with the exception of end- and side-scrapers. Arrowheads,
pieces in the higher translucency classes. Again, the JKV too, at JKV were mainly made from blades, although their
sample does not resemble the Rijckholt material, even if small size often makes it difficult to determine the type of
only the raw data are considered. The correspondence blank. Most retouched tools show a direct steep retouch,
with the data from Rullen, Banholt and Hoogbos is much with only arrowheads sometimes displaying bifacial or in-
better, and improves when regarding the enhanced data; verse flat retouch.
Rodebos, on the other hand, seems a less likely source, be- The cores and hammerstones discarded at JKV are quite
cause flints with a low or medium translucency are scarce small (Table 5).The blades are rather stocky; three quarters
at that location. (76.5%) of them have a length/width ratio between 1:2 and
Thus, it seems highly probable that the majority of JKV 1:3 (and 48.1 % between 1:2 and 1:2.5).
raw material was collected at the Banholtergrub (or at one
or more as yet unknown extraction sites, where the flints 3. Diachronic observations
were embedded in similar conditions), with Hoogbos and
Rodebos as possibilities of secondary importance. In all habitation phases, the percentage of unmodified
The flints of the Bandkeramik sites at Elsloo and Beek- flakes at JKV is very high (between 67.4% and 77.3%),
Kerkeveld display the same characteristics (observation as is the percentage of unmodified flakes with cortex (be-
by the present author). Given the mention of transparent tween 34 and 39 %). Nevertheless, the assemblage displays
reddish-brown zones as typical for the so-called Rijck- important diachronic variations (Table 6). It is not easy to
holt flint encountered elsewhere in the Bandkeramik assess these figures looking only at percentages, firstly be-
world (e.g. Deutmann 1997, Lhr et al.1977, Zimmer- cause of their interdependence, and secondly because of
mann 1988, 606), it seems plausible that this material the enormous difference in frequency between unmodified
mainly originated from Banholt as well. It even reached as flakes and cores. Therefore, the relative proportions of
far east as Friedberg-Bruchenbrcken in Hesse, where at main artefact categories are presented as well (Table 7).
least two artefacts made of the Banholt variety were iden-
tified among the Oldest Bandkeramik material by A. L. In the first habitation phase (E1) the proportion of flakes
Fischer and the present author, using the authors reference to cores is considerably lower than in the other phases, as
collection (Fischer 2005, De Grooth in prep.). is the proportion of flakes to blades plus tools. Two main
interpretations are possible for this phenomenon. In the
2. The process of flint working at first, the phase E1 assemblages have too few flakes for
Geleen Janskamperveld every core, and the situation in the other phases is regarded
as standard. If, alternatively, the phase E1 situation is re-
The chane opratoire (Pelegrin et al. 1988) chosen by garded as standard, the other phases would have too many
JKVs inhabitants to transform their raw material into man- flakes, i.e. too few cores.
ufactured products strongly resembles the one practised at In the first case, during phase E1 flakes would have been
other regional Bandkeramik sites, such as Beek-Kerkeveld underrepresented at JKV because the preparation of (some
20
Types Early habitation All material
N % N %
Arrowheads 42 8.6 74 8.3
Borers 28 5.7 47 5.3
End-scrapers 144 29.5 264 29.7
Side-scrapers 13 2.7 23 2.6
Truncated blades 41 8.4 62 7.0
Retouched blades 80 16.4 179 20.1
Sickle blades 80 16.4 131 14.7
Retouched flakes 25 5.1 44 4.9
Splintered pieces 17 3.5 37 4.2
(micro) Burins 4 0.8 5 0.6

Notches 8 1.6 11 1.2
Denticulates 2 0.4 4 0.5
Microliths 3 0.6 3 0.3
Quartiers dorange 0 0.0 1 0.1
Non-sickle gloss 1 0.2 5 0.6
489 100.0 890 100.1
Table. 4: JKV, the main tool types.
2
Surface mm
Tablets Length Width
61 47 3039
Blade cores Platform L Platform W
46 40 1955
Blades/blade tools Width Thickness
18 5 90
Table. 5: JKV, estimate of the amounts of blades produced (median dimensions in mm).
E1 % E2 % E3 % E4 % All %
Cores 13 1.3 11 0.8 18 1.0 5 0.7 47 1.0
Flakes 671 68.7 1028 75.4 1364 75.0 475 67.4 3538 72.7
Blades 158 16.2 173 12.7 223 12.3 114 16.2 668 13.7
Tools 102 10.4 129 9.5 182 10.0 96 13.6 509 10.5
Table. 6: JKV, diachronic variability of main artefact types through time.
E1 E2 E3 E4
Flakes : Blades+Tools 2.6 3.4 3.4 2.3

Flakes : Cores 51.6 93.5 75.8 95.0
Blades+Tools : Cores 20.0 27.5 22.5 42.0
Blades : Tools 1.6 1.3 1.2 1.2
Flakes : Tools 6.6 8.0 7.5 4.9
N 977 1363 1819 705
Table. 7: JKV, proportions of main artefact categories through time.
21
of the) cores had taken place elsewhere, e.g. at the extraction the low ratio of blades plus tools to cores); the number of
sites, whilst during the other phases unprepared cores were exported cores may have been somewhat lower than in the
brought into the settlement. As this practice would not influ- preceding phase.
ence the intensity of blade and tool production at the site, one E4: the number of exported cores increased again, but export
would expect the proportions of blades and tools to cores to mainly took place after initial blade production (hence the
be similar for all phases. This clearly not being the case, the high ratio of blades plus tools to cores).
alternative interpretation seems to fit better. As it is known A corroboration of this interpretation is provided by a com-
that flints of the Lanaye type circulated widely throughout parison of the size of blade cores and rejuvenation tablets
the Bandkeramik world, the most plausible scenario would discarded at JKV. The differences in surface of the striking
be based on the assumption that production and use of cores platforms and of the complete rejuvenation tablets, allow for
was a local affair during phase E1, but that in the other an estimate of the amount of blades to be made (Fig. 4). The
phases some of the cores prepared at the settlement were not median platform surface of cores is 1955 mm; the complete
discarded there, but exported. Moreover, the fluctuations in tablets have a median surface of 3039 mm. Thus, a median
the ratio of cores to blades plus tools, and of flakes to blades surface of 1084 mm was removed during one stage of blade
plus tools indicate that this export may have occurred at dif- production. Given a median surface (i.e. width x thickness)
ferent stages in the reduction process as well. It is not easy of 90 mm per blade, about 12 blades would have resulted
to interpret these figures in more detail, partly because they (Table 5). Presuming that the same number of blades was
may result from several different strategies, partly because produced before core rejuvenation, every core might have
of the many parameters involved. A tentative interpretation yielded some 25 blades. If most cores discarded at JKV orig-
would be: inally produced blades, the proportions of cores to blades
E1: local production; no cores, but some (un)modified blades plus tools for the phases E2 and E3 would correspond quite
were exported (hence the relatively low proportion of blades nicely to this estimate, supporting the idea that some nodules
and tools in relation to both flakes and cores). were prepared as cores at JKV, but exported before blades
E2: local production, but some of the cores were exported were struck from them. In E1 the number of blades/tools
(hence the higher ratio of flakes to cores), possibly some of is somewhat lower, and in E4 much higher than would be
them after preparation, some after initial blade production expected. Thus, during phase E1 some (un)modified blades
(resulting in a higher ratio of blades plus tools to cores). may have been exported, whereas in E4 cores were used for
E3: a part of the cores were exported after preparation (given blade production at JKV before leaving the settlement.
Blades
Original 18 x 5 mm
nodules
Exhausted
Tablets Cores
61 x 47 mm 46 x 40 mm
Fig. 4: JKV, relationships between dimensions of blade cores, rejuvenation tablets and blades.
22
4. Inter-site comparisons pieces with cortex, as well as a smaller amount of unre-

touched blades and tools than is found at neighbouring
To assess the value of this interpretation, one would like sites, Langweiler 8 is often interpreted as having played a
to compare the JKV-data with those from the settlements central role in the production and distribution of flint cores
that could have received its flints, especially sites further and blanks (Kegler-Graiewski and Zimmermann 2003,
to the east that used important quantities of Western La- Zimmermann 1995, 2002). In this early period, however,
naye flints, as well as from contemporary settlements in the differences with the presumed consumer settlements of
the Graetheide region, located at approximately the same the Middle Merzbach valley are not really clear cut. Ob-
distance from the extraction sites. viously, it took time for Langweiler 8 to establish itself
Unfortunately, at the moment is it impossible to synchro- as a regional flint redistribution centre. Weisweiler 17 is
nize the JKV chronology with the four micro habitation also described as a flint producing central place (Kegler-
phases that comprise the Early LBK period at Elsloo (cf. Graiewksi and Zimmermann 2003; Zimmermann 2004).
Van de Velde 1979). On the other hand, I am fairly confi- As this site provided only 214 datable flint artefacts (of
dent that, despite the differences in the analytic techniques which 126 belong to the Early LBK), it is difficult to as-
applied, one may equate the JKV habitation phases as de- sess this claim.
fined by Van de Velde (2007) with the House Generations Therefore, the comparison had to be performed at a more
(HG) of the Rhineland, as they are based on the analysis general level, concentrating on JKV, Langweiler 8 and
of quite similar decorated pottery. Both phase JKV-E1 and Elsloo, and looking at the Flomborn period as a whole. Of
House Generation I belong to the beginning of Modder- course, the data may not be used directly to infer differ-
mans (1970) phase Ib, JKVs generation E4 is situated ences in the ways flints were worked, without taking into
somewhere during Modderman Ic, as are House Genera- account differences in excavation and sampling methods
tions IV and V, with HG VI and VII representing Modder- and in post-depositional processes. The three sites share
man Id, when JKV had been abandoned (cf. D. Mischka two handicaps: firstly, the topsoil was removed mechani-
2004, Abb. 9 ). cally, causing the loss of a considerable amount of material
The overall number of flint artefacts at LW 8 is consider- (mainly tools and blades), as witnessed by the excavation
able: 9532 (Zimmermann 1988, Abb. 576), of which 7554 at Sittard. There, the manually removed topsoil contained
are assignable to dated features (D. Mischka 2004, Abb. 55% of the tools, 56% of the cores and hammerstones,
15). The amount of flint recovered at Langweiler 8 and but only 37% of the waste flakes and blades (Modderman
the other settlements on the Aldenhovener Platte during 1958/59, 113). Secondly: the pits contents were not sys-
the Early LBK (comprising the first seven House Genera- tematically sieved, resulting in the possible under-repre-
tions), however, is so small as to preclude a meaningful sentation of smaller artefacts (Gronenborn 1997). Elsloo
comparison on the level of House Generations (Table 8). and Langweiler 8 were investigated during rescue excava-
At best one may observe that the percentage of unmodified tions, often under extreme time stress (Zimmermann 1987,
flakes at JKV is consistently higher than that at Langweiler 636), whilst JKV, as a student training project, could be
8. Given the high percentage of unmodified flakes and of investigated in a (somewhat) more leisurely way. Moreo-
JKV E1 E2 E3 E4
N 977 1363 1819 705
% flakes 68.7 75.4 75.0 67.4
LW8 I II III IV V VI VII
N 49 230 209 553 774 188 262
% flakes 59.1 61.7 67.5 62.0 64.2 64.9 64.5
LW2, LW9, LB7
N 38 116 1214 453 115 110 1248
% flakes 57.9 64.7 51.8 52.8 69.6 60.9 59.1
Table. 8: Comparison of the number of flint artefacts and of unmodified flakes at JKV, Langweiler 8, and other Early LBK settlements in the Middle
Merzbach valley (Langweiler 2, Langweiler 9, Laurensberg 7). Merzbach valley data according to D. Mischka 2004, Abb. 15.
23
ver, the preservation at JKV is considered to be better than These observations may be interpreted in the following
at the other sites (Louwe Kooijmans et al. 2003). Both at way:
Elsloo (observation by the present author) and at Lang- The three sites under consideration all have very high per-
weiler 8 (Zimmermann 1988, 635) only c. 3% of the un- centages of unmodified flakes, and thus would qualify as
modified flakes have a length under 15 mm, against 18% sites where flint was worked locally. In terms of the mod-
at JKV. The mean weight of unmodified flakes also differs: els developed by Zimmerman (1995) to study exchange
5.8 g. (or 7.0 g when the chips are excluded) for JKV vs. mechanisms on the Aldenhovener Platte, the difference
9.7 g. for Langweiler 8 (no data on weight are available for in the percentage of unmodified flakes between JKV and
Elsloo). These differences, however, could partly have to Elsloo on the one hand and Langweiler 8 on the other
do with the intensity of flint working per se as well: where hand, would indicate that the earlier stages of core reduc-
more preparation takes place, more chips and smallish tion were better represented at the former settlements. In
flakes are produced. To neutralize the possible bias caused other words: some of the cores arriving at Langweiler 8
by JKV being more carefully excavated, its data will be had been prepared elsewhere.
presented with and without chips (Table 9). From this data As JKV and Elsloo both have a very high percentage of
a number of observations may be derived: flakes, the low index of flakes to cores at Elsloo cannot
-The percentage of unmodified flakes at both JKV and be explained by assuming that core preparation was not
Elsloo is considerably higher than that at LW8. performed locally. Instead, I think that the Elsloo cores re-
-The percentage of cores is lowest at JKV, and highest at mained in the settlement, whilst part of the JKV cores were
Elsloo, with LW8 in an intermediate position. exported. Instead, Elsloo was an exporter of blades; wit-
-The ratio of flakes to cores at JKV is very much higher ness its low ratio of blades plus tools to cores. This is cor-
than at both LW8 and Elsloo. roborated by the high index of flakes to blades plus tools.
-At JKV there are 14 blades and 11 tools for every core; If Langweiler 8 was a receiver of JKV cores, one would
at LW8 7.7 blades and 6.6 tools, and at Elsloo only 3.9 expect the flake/core index at this site to have been lower
blades and 2.4 tools. than that of Elsloo. The relatively high index can be under-
-At JKV seven (or six) flakes are present for every tool; stood, however, when taking into account that Langweiler
at LW8 four and at Elsloo 10.5. 8 in its turn was an exporter of further reduced (and re-
-At JKV there are 5.4 (or 4.4) flakes for every blade, at juvenated) cores (cf. Graiewski and Zimmermann 2003).
LW8 3.6 and at Elsloo 6.4 Moreover, if the blades missing at Elsloo were in part
JKV JKV LW8 Elsloo

(no chips)
N 4864 4241 1351 3515
% flakes 72.7 68.7 63.2 76.0
% cores 1.0 1.1 2.3 2.9
% blades plus tools 24.2 27.8 32.8 18.9
Flakes : Cores 75.3 62.0 27.5 26.4
Blades: Cores 14.2 14.2 7.7 3.9
Tools: Cores 10.8 10.8 6.6 2. 4
Flakes : Tools 7.0 5.7 4.2 10.5
Flakes: Blades 5.3 4.4 3.6 6.4
Table. 9: Comparison of the intensity of flint working at JKV, Langweiler 8 and Elsloo during the Flomborn period (Langweiler 8: Zimmermann 1988,
Abb. 596. Elsloo: De Grooth 1987).
24
JKV LW 8 Early Ald. Platte, Hambach 8

Early (I-VII) general
Tablets N 31 8 163 48
Mean L (mm) 60.8 54.3 56.2 52.0
Mean W (mm) 47.7 38.5 41.9 37.8
Cores N 44 18 628 28
Mean L (mm) 53.7 51.8 56.8 50.6
Mean W (mm) 44.9 41.2 43.7 34.9
Mean Th 34.0 28.3 32.1 22.7
Table. 10: Comparison of dimensions of rejuvenation tablets and exhausted cores. LW 8 Early: data made available by A. Zimmermann, Cologne;
Aldenhovener Platte and Hambach 8 according to Hohmeyer 1997, Taf. 54, 59, 71, 72.
transferred to LW8, they help account for the low ratio of tion sites, they used different procurement strategies, and
flakes to blades and tools, and the high ratio of blades plus Langweiler 8 was in part dependent on cores and blades
tools to cores. from the Graetheide settlements (Fig. 5).
-A nice corroboration of the idea that JKV exported part At first sight, this variability in procurement and exchange
of its cores to Langweiler 8 is provided by the size of the strategies is surprising, as Bandkeramik long-distance ex-
core rejuvenation tablets (Table 10). Whereas the exhaust- change networks generally are thought to be based on long-
ed cores of JKV and Langweiler 8 have similar sizes, the standing, stable kinship ties, ties carefully maintained from
tablets discarded at Langweiler 8 are considerably smaller, one generation to the next (e.g. Krahn-Schigiol 2005, Lech
as though they have served to rejuvenate cores at a later 2003). A general overview of Flomborn-period population
stage of the reduction sequence. The data for the Alden- dynamics may provide an explanation.
hovener Platte as a whole and for Hambach 8 (some 10
km to the north-east) confirm the trend, especially as they
comprise material from both the Older and the Younger
JKV
LBK (when cores tend to be larger).
prep are d/reduce d
Core s
The debris recovered at the Banholt extraction site makes

Elsloo
the pattern even more complicated, as it yielded not only
Blades
exhausted blade cores, but also quite a number of rejuve-
nation tablets (collected by the present author). Apart from
LW 8
large polyhedral blade cores such as those published by
Brounen and Peeters (2009/2001) which recall the cores
worked at Beek-Kerkeveld and other Younger LBK sites, re duced
Core s
the extraction activities are undated. However, smaller
cores, closely resembling the JKV material, are present as Nodul es
well. Blades
Banho lt
Therefore it seems plausible that during some of the time, c.s.
some of the material was brought into some of the set-
tlements under consideration in the shape of blades pro-
duced at the extraction site, as was the case for instance in
Fig. 5: Possible procurement and exchange strategies practised during
the Guboden area of South-eastern Bavaria (De Grooth the Early LBK in the Rhine-Meuse area. Thick continuous line: direct
procurement of nodules. Thick dotted line: transport in the framework of
2003a). Thus, although the three settlements were locat-
down-the-line exchange. Thin dotted line: transport after direct procure-
ed at approximately the same distance from the extrac- ment and production of blanks at extraction sites.
25
Radiocarbon and ceramic dates alike indicate that JKV was exchange routines that were shown to exist between JKV
a first generation settlement, as were Geleen-Kluis, Sitta- (and other early Graetheide settlements) and the Rhenish
rd, Elsloo and Stein (Van de Velde et al. in press). In other settlements. On the contrary, the major incentive to main-
words, settlement west of the river Rhine was not a gradual, tain alliances with eastern neighbours and kin may have
tentative step-by-step process, but started with a great leap been the western settlers need for a continuous supply of
westward, followed by filling-in of the areas in between amphibolite and basalt adzes.
during the next generations. Several sites in this Hinter- On the one hand, the pioneer situation outlined above
land were settled in the same early stage as well, the best- would induce people to cherish and maintain kinship ties.
studied being Langweiler 8 on the Aldenhovener Platte. On the other hand, this was a time of immense change in
Mnchs (1999, 2005; cf. D. Mischka 2004, Zimmermann settlement and habitation, where one could not depend on
2002) recent re-analysis of this areas decorated ceramics traditional, fixed exchange networks alone, but had to be
enabled her to give an exemplary insight in the dynamics flexible and opportunistic. Yesteryears trusted exchange
of settling in the seven house generations of the Flomborn partner and his family today may have moved on, or may
period for this settlement and its neighbourhood, compris- be bound by obligations to other relations. In such an un-
ing an area of some 25 square kilometres (Table. 11). In- stable situation, differentiated procurement and exchange
habitation began at Langweiler 8, with four contemporary strategies were called for, even on a micro-regional scale.
Therefore, the fluctuations in the export of flint found at
House generation N settlements N Houses JKV, in my view, are perfectly compatible with the alterna-
I 1 4 tion of expansion and consolidation and the multiple net-
II 7 14 works visible in the Rhineland.
III 8 20
IV 9 23
V 8 24
VI 9 26 Marjorie E. Th. de Grooth
VII 9 29
Aspelweg 49
D-53902 Bad Mnstereifel-Eicherscheid
Table. 11: Summary of settlement dynamics along the middle Merzbach-
valley and its surroundings during the Flomborn period (after Mnch Deutschland
2005). e-mail: grooth@t-online.de
houses. In the next generation, not only did the number of

houses there increase to 7, but at least six new settlements References
started, mostly inhabited by a single household. In the next
house generations, some of these too grew in size, others Ahler, S.A. 1983. Heat treatment of Knife River flint.
remained single farmsteads, or were abandoned. In the last Lithic Technology 12, 1-8.
Flomborn generation, almost 30 houses are known to have
been inhabited. Similar dynamics are documented for other Allard, P. 2005. Lindustrie lithique des populations
parts of the Rhineland (Zimmermann et al. 2004, Claen rubanes du Nord-Est de la France et de la Belgique. In-
2006), even Erkelenz-Kckhoven, located some 23 km to ternationale Archologie 86. Rahden/Westphal: Verlag
the north, on the northern fringe of the Rhenish loess area, Marie Leidorf GmbH.
was first settled in House Generation III (Lehmann 2004).
At all of these sites, Lanaye flints, in part definitely of the Bakels, C.C. 1978. Four Linearbandkeramik Settlements and
Banholt variety, played an important role. their Environment. Analecta Praehistorica Leidensia 11.
Given the fact that all settlements under consideration
were located within one days walking distance from the Bohmers, A. and Bruijn, A. 1958. Statistische und gra-
extraction sites, and could acquire alternative raw material phische Methoden zur Untersuchung von Flintkomplexen
of reasonable quality in local river gravels (Weiner 1997), IV. Das lithische Material aus den bandkeramischen Sied-
there was no intrinsic need for the establishment of the lungen in den Niederlanden. Palaeohistoria 6/7, 183-213.
26
Brounen, F.T.S. and Peeters, H. 2000/2001. Vroeg-neoli- Gehlen, B. and Schn, W. (in press). Jngere Band-
thische vuursteenwinning en -bewerking in de Banholter- keramik frhes Mittelneolithikum Rssen im
grubbe (Banholt, gem. Margraten). Archeologie 10, 133- rheinischen Braunkohlenrevier: Steinartefakte als Spiegel
150. einer sich verndernden Welt, Rheinische Ausgrabungen.
Cahen, D. 1984. Technologie du dbitage laminaire. In M. Giot, D., Mallet, N. and Millet, D. 1986. Les silex de
Otte (ed.) Les fouilles de la Place Saint-Lambert Lige la rgion du Grand-Pressigny (Indre-et-Loire). Re-
1, Lige: ERAUL 18, 171-199. cherche gologique et analyse ptrographique. Revue
Archologique du Centre de la France 25, 21-36.
Cahen, D., Caspar, J.P. and Otte, M. 1986. Les industries
lithiques danubiennes de Belgique. Lige, ERAUL 21. Gronenborn, D. 1997. Silexartefakte der ltestband-
keramischen Kultur. Bonn: Universittsforschungen zur
Claen, E. 2006. Die bandkeramische Siedlungsgruppe Prhistorischen Archologie 37.
bei Knigshoven. Unpublished PhD thesis, Cologne.
Grooth, M.E.Th. de, 1987. The Organisation of Flint Tool
Deutmann, K.H. 1997. Feuersteinmaterial. In U. Boe- Manufacture in the Dutch Bandkeramik. Analecta Prae-
licke, K.H. Deutmann, J. Lning, F.W.V. Schmidt and P. historica Leidensia 20, 27-52.
Stehli, Der bandkeramische Siedlungsplatz Aldenhoven 3,
Kr. Dren, in J. Lning Hrsg., Studien zur neolithischen Grooth, M.E.Th. de 1998. Archeologische beschrijvin-
Besiedlung der Aldenhovener Platte und ihrer Umgebung, gen van Ryckholt-vuursteen. In P.C.M. Rademakers (ed.),
Rheinische Ausgrabungen 43. Kln/Bonn, 83-125. De prehistorische Vuursteenmijnen van Ryckholt-St.
Geertruid, Maastricht, 160-163.
Felder, P.J. and Felder, W.M. 1998. De geologie van de
omgeving van het prehistorisch mijnveld te Ryckholt-St. Grooth, M.E.Th. de 2003a. Dealing with Bandkeramik
Geertruid. In P.C.M. Rademakers (ed.), De prehistorische cherts. Procurement strategies in south-eastern Bavaria.
Vuursteenmijnen van Ryckholt-St. Geertruid, Maastricht, In L. Burnez-Lanotte (ed.). Production and Management
111-137. of Lithic Materials in the European Linearbandkeramik.
Acts of the XIVth UISPP Congress, University of Lige,
Felder, P.J., Rademakers, P.C.M. and de Grooth, M.E.Th. Belgium 2-8 September 2001, BAR International series
(eds.) 1998. Excavations of Prehistoric Flint Mines at 1200. Oxford, 37-44.
Rijckholt-St. Geertruid (Limburg, The Netherlands).
Archologische Berichte 12. Bonn. Grooth, M.E.Th. de 2003b. They do things differently
there. Flint working at the Early Bandkeramik settle-
Felder, W.M. 1998. Overzicht van de prehistorische ment of Geleen-Janskamperveld (The Netherlands). In J.
vuursteenexploitaties binnen het Krijtgebied tussen Eckert, U. Eisenhauer and A. Zimmermann (eds.), Archol-
Aken-Heerlen-Luik-Maastricht en Tongeren. In P.C.M. ogische Perpektiven; Analysen und Interpretationen im
Rademakers (ed.), De prehistorische Vuursteenmijnen van Wandel; Festschrift fr Jens Lning zum 65. Geburt-
Ryckholt-St. Geertruid, Maastricht, 169-193. stag, Internationale Archologie: Studia honoria Bd. 20.
Rahden/Westf, 401-406.
Felder, W.M. and Bosch, P.W. 2000. Krijt van Zuid-Lim-
burg. Geologie van Nederland deel 5. Delft/Utrecht. Grooth, M.E.Th. de (in prep). Distinguishing Upper Cre-
taceous flint types exploited during the Neolithic in the
Fischer, A.-L. 2005. Die Silexartefakte der bandkera- region between Maastricht, Tongeren, Lige and Aachen.
mischen Siedlung Bruchenbrcken, Stadt Friedberg/Hes- In J. Meurers-Balke and W. Schn (eds.) Gedenkschrift fr
sen. Unpublished MA thesis. Cologne. Jrgen Hoika. Bonn: Archologische Berichte.
27
Grooth, M.E.Th. de 2007. Flint working in the Bandkeramik Prhistorique, Bruxelles 1-8 septembre 1935, 348-357.
settlement. In P. van de Velde et al. Excavations at Geleen-
Janskamperveld, Analecta Praehistorica Leidensia 39. Louwe Kooijmans, L.P., van de Velde, P. and Kamermans,
H. 2003. The early Bandkeramik settlement of Geleen-
Kegler-Graiewski, N. and Zimmermann, A. 2003. Ex- Janskamperveld: Its intrasite structure and dynamics.
change systems of stone artefacts in the European Neolithic. In J. Eckert, U. Eisenhauer and A. Zimmermann (eds.),
In L. Burnez-Lanotte (ed.), Production and Management of Archologische Perpektiven; Analysen und Interpreta-
Lithic Materials in the European Linearbandkeramik. Acts tionen im Wandel; Festschrift fr Jens Lning zum 65.
of the XIVth UISPP Congress, University of Lige, Bel- Geburtstag, Internationale Archologie. Studia honoria
gium 2-8 September 2001, BAR International Series 1200. Bd. 20, Rahden/Westf, 373-398.
Oxford, 31-35.
McDonnell, R.D, Kars, H. and Jansen, J.B.H. 1997. Pe-
Krahn-Schigiol, Chr. 2005. Regionale und berregionale trography and Geochemistry of Flint from Six Neolithic
Kommunikationsnetzwerke der bandkeramischen Sied- Sources in Southern Limburg (The Netherlands) and
lungen auf der Aldenhovener Platte. In J. Lning, Chr. Northern Belgium. In: A. Ramos-Milln and M.A. Bustillo
Frirdich, A. Zimmermann (eds.), Die Bandkeramik im (eds.), Siliceous Rocks and Culture, Granada, 371-385.
21. Jahrhundert. Symposium in der Abtei Brauweiler bei
Kln vom 16.9-19.09.2002. (Internationale Archologie: Mischka, D. 2004. Zentraler Ort oder Nebensiedlung?
Arbeitsgemeinschaft, Symposium, Tagung, Kongress; Bd Die Feinchronologie der Grundformspektren des band-
7). Rahden/Westf., Posterbeilage 1. keramischen Fundplatzes Kckhoven im Vergleich. In H.
Koschik (ed.), Der bandkeramische Siedlungsplatz von
Lech, J. 2003. Mining and siliceous rock supply to the dan- Erkelenz-Kckhoven, Kreis Heinsberg I. Archologie,
ubian early farming communities (LBK) in eastern central Rheinische Ausgrabungen 54. Mainz, 537-594.
Europe: a second approach. In L. Burnez-Lanotte (ed.).
Production and Management of Lithic Materials in the Eu- Modderman, P.J.R., 1958/1959. Die bandkeramische Sied-
ropean Linearbandkeramik. Acts of the XIVth UISPP Con- lung von Sittard. Palaeohistoria 6-7, 33-120.
gress, University of Lige, Belgium 2-8 September 2001,
BAR International series 1200. Oxford, 19-30. Modderman, P.J.R., 1970. Linearbandkeramik aus Elsloo
und Stein. Analecta Praehistorica Leidensia 3.
Lehmann, J. 2004. Die Keramik und Befunde des band-
keramischen Siedlungsplatzes Erkelenz-Kckhoven, Mnch, U. 1999. Zur Siedlungsstruktur der Flombornzeit
Kreis Heinsberg (Grabungskampagnen 1989-1994). In H. auf der Aldenhovener Platte. Unpublished MA thesis, Co-
Koschik (ed.), Der bandkeramische Siedlungsplatz von logne.
Erkelenz-Kckhoven, Kreis Heinsberg I. Archologie,
1-364. Rheinische Ausgrabungen 54. Mainz. Mnch, U. 2005. Zur Siedlungsstruktur der Flombornzeit
auf der Aldenhovener Platte. In J. Lning, Chr. Frirdich,
Lhr, H., Zimmermann, A. and Hahn, J. 1977. Feuerstei- and A. Zimmermann (eds.), Die Bandkeramik im 21. Jahr-
nartefakte. In R. Kuper, H. Lhr, J. Lning and A. Zim- hundert. Symposium in der Abtei Brauweiler bei Kln vom
mermann, Der bandkeramische Siedlungsplatz Langweiler 16.9-19.09.2002. (Internationale Archologie: Arbeitsge-
9, Ge. Aldenhoven, Kr. Dren, Beitrge zur neolithischen meinschaft, Symposium, Tagung, Kongress; Bd 7). Rah-
Besiedlung der Aldenhovener Platte 2. Rheinische Aus- den/Westf, Posterbeilage 2.
grabungen 18. Bonn, 131-266.
Pelegrin, J., Karlin, C. and Bodu, P. 1988. Chanes opra-
Louis, M. 1936. Les ateliers nolithiques du Hoog toires: un outil pour le prhistorien. In J. Tixier (ed.),
Bosch (Fouron-le-Comte, province de Lige), XVIe Technologie prhistorique, Notes et monographies tech-
Congrs International dAnthropologie et dArchologie niques du CRA 24. Paris, 37-53.
28
Velde, P. van de, et al. (2007). Excavations at Geleen- ancien et moyen en France et dans les rgions limitrophes.
Janskamperveld. Analecta Praehistorica Leidensia 39. Actes de la Xme Session de lEAA, Lyon Septembre 2004,
BAR International series 1494. Oxford: 80-83.
Verhart, L.B.M., 2000. Times fade away. The neolithiza-
tion of the southern Netherlands in an anthropological and Zimmermann, A., Richter, J., Frank, Th. and Wendt, K.P.
geographical perspective. Leiden: Archaeological Studies 2004. Landschaftsarchologie II berlegungen zu Prin-
Leiden University 6. zipien einer Landschaftsarchologie. Bericht der Rmisch-
Germanischen Kommision 85, 39-96.
Vermeersch, P.M., Chow, J., Creemers, G., Masson-Loodts,
I., Groenendijk, A.J and DeBie, M. 2005. Neolithische
vuursteenontginning op de site van Rullen (Voeren, prov. Acknowledgements
Limburg). In I. In t Veld and W. De Clercq (eds), Een lijn
door het landschap. Archeologie en het vTn-project 1997- Many thanks to Thomas Frank and Andreas Zimmermann
1998, Archeologie in Vlaanderen 5. Brussel, 313-328. (Cologne) for Figure 1, and for providing unpublished data
on the Langweiler 8 tablets; to my Leiden University col-
Warrimont, J.P. de and Groenendijk, A.J. 1993. 100 jaar leagues Annelou van Gijn, Leendert Louwe Kooijmans,
Rullen vuursteen: een kleurrijke vuursteensoort nader Piet van de Velde and Ivo van Wijk for their help in dealing
bekeken, Archeologie in Limburg 57, 37-46. with JKVs flints; to Rengert Elburg, Jean-Pierre de War-
rimont and Kim Groenendijk, Joep Franzen, Henk and Eva
Weiner, J. 1997. Die Maasschotter der Niederrheinischen Kars, Leo Verhart, Pierre Vermeersch, and Werner Felder
Bucht als Feuersteinlieferant fr die bandkeramischen and his fellow members of the Working Group on Pre-
Siedlungspltze Langweiler 8 und Langweiler 9. In J. historic Flint Mining for stimulating discussions on flint
Lning, (ed.), Studien zur neolithischen Besiedlung der characterization; and last, but certainly not least, to Paul
Aldenhovener Platte und ihrer Umgebung (Beitrge zur Wheeler for improving the English text.
neolithischen Besiedlung der Aldenhovener Platte 6),
Rheinische Ausgrabungen 43. Kln/Bonn, 599-646.
Zimmermann, A. 1988. Steinmaterial. In U. Boelicke, D.

von Brandt, J. Lning, P. Stehli and A. Zimmermann, Der
bandkeramische Siedlungsplatz Langweiler 8, Gemeinde
Aldenhoven, Kreis Dren (Beitrge zur neolithischen
Besiedlung der Aldenhovener Platte III), Rheinische
Ausgrabungen 28. Bonn, 569-787.
Zimmermann, A. 1995. Austauschsysteme von Silexarte-

fakten in der Bandkeramik Mitteleuropas. Bonn: Univer-
sittsforschungen zur prhistorischen Archologie 26.
Zimmermann, A. 2002. Landschaftsarchologie I: Die

Bandkeramik auf der Aldenhovener Platte. Bericht der
Rmisch-Germanischen Kommission 83, 17-39.
Zimmermann, A. 2006. Production and demand of flint

artefacts in the Bandkeramik of western Germany. In P. Al-
lard, F. Bostyn and A. Zimmermann (eds.), Contribution
des matriaux lithiques dans la chronologie du Nolithique
29
30
Pierre Allard and Laurence Burnez-Lanotte: An economy of surplus production in the Early Neolithic of Hesbaye (Belgium):
Bandkeramik blade debitage at Verlaine Petit Paradis
An economy of surplus production in the Early

Neolithic of Hesbaye (Belgium): Bandkeramik
blade debitage at Verlaine Petit Paradis
Pierre Allard, Laurence Burnez-Lanotte
Abstract: Research conducted on a lithic assemblage from the Early Neolithic at Hesbaye led us to explore the phenom-
enon of flint blade blank production with an original management strategy. This paper addresses a particular aspect of this
complex economic and social phenomenon, which can be summarized by the following question: how can we qualify the
production of blades in this precise case, and can we formalize and demonstrate the hypothesis of an overproduction, as
we have already proposed. The problem of quantifying debitage products in concentrations, houses, and entire villages
can be addressed at Verlaine based on the analyses that we propose here. The question is clearly not to quantify the exact
number of blades produced, such an objective being inaccessible in our current state of research, but rather to formulate
acceptable estimates. The method to evaluate the total number of blades requires several steps. In this paper, we shall
present only the first attempts, based upon a few cases. The test we propose is based on the integration of three methods
applied to the artefacts of a debitage concentration and detritic waste found within the same structure.
Keywords: Neolithic, flint, Bandkeramik, overproduction.
Introduction
Research conducted on a lithic assemblage from the Early attempt to define the notion of an economy of lithic sur-
Neolithic at Hesbaye led us to explore the phenomenon of plus production, which corresponds to an original type of
a flint blade blank production with an original manage- blank production management.
ment strategy. This paper addresses a particular aspect of
this complex economic and social phenomenon, which 1. The archaeological context of the
can be summarized by the following question: how can production
we qualify the production of blades in this precise case,
and can we formalize and demonstrate the hypothesis of The site of Verlaine is located 15 km from Lige, Belgium
an overproduction such as we have already proposed (Fig. 1). From 1996 to 2002, L. Burnez-Lanotte, Univer-
(Burnez-Lanotte and Allard 1998, 2003). What method sity of Namur, conducted an exhaustive excavation of the
can we use to estimate the quantity of blades produced? site, confirming the presence of a Bandkeramik (or Lin-
This implies estimating the number of blades produced ear Pottery Culture) settlement extending over more than
by the best documented laminar productions of the settle- three hectares. The occupation includes several houses,
ment, and establishing a quantitative relationship between 100 pits and nearly 22 flint debitage concentrations. In
this calculation and the number of blades created by this 2001, four debitage concentrations were discovered in
production and used in the village. On this basis, we shall the lateral pits of two joined houses. The concentration of
31
Fig. 1: Verlaine Petit Paradis, Belgium. Plan of the site 19962001; black stars indicate the debitage concentrations (Burnez-Lanotte, Allard 2003).
32
structure 130 was perfectly preserved in the bottom of the procedure corresponds to a successive rhythm that begins
pit (Burnez-Lanotte and Allard op. cit.). A detailed techno- on one side of the core and crosses the entire debitage sur-
logical study of this concentration, including exhaustive face toward the other side (rhythm 1-2-3 or 3-2-1). The
refitting, showed the particularities of the Verlaine artefact first procedure, in association with the calibration of the
assemblage, whose debitage objectives seem to be related products, is indicative of standardized and well controlled
to an economic phenomenon of flint blade production in series (Binder and Gassin 1998).
larger than usual quantities. Our analyses show that these - A high skill level. Though simple, this debitage of short
products were exported and integrated within a circulation blades by indirect percussion was performed by good
network covering the northern part of the Bandkeramik knappers. The absence, or small amount of butt prepara-
zone in Western Europe (Burnez-Lanotte and Allard op. tion, the large size of the core tablets, the complex prepara-
cit., Allard 2005, 2007). tions, as well as multiple opposed debitage sequences, are
all signs of a high general knapping standard.
2. Blade debitage at Verlaine: the six - Reduction sequences adapted to an intended production,
characteristic elements of this tech- aiming at maximal production. The refits of concentration
nical production
130 show that the management of volumes was oriented
uniquely toward the production of standardized blades, 8
We have already presented elsewhere the particularities of to 12 cm long. The debitage of blocks more than 20 cm
blade production at Verlaine (Allard and Burnez-Lanotte long is particularly significant; this could have allowed for
2007). They include a number of quantitative and qualita- the production of longer blades, but we observe instead
tive factors that can be summarized as follows: the successive debitage of multiple blade surfaces with
- The quantity of lithic artefacts at Verlaine is higher than changes in the volumes exploited and sequences of partial
at other sites currently excavated. In a preliminary study of or total re-preparation of the block, to insure an optimal
13 pits (including 6 lithic concentrations), we have already production of standardized blades (Allard 2007).
recorded nearly 60 000 objects including more than 1 000
blade cores. All the structures in the village contain lami- These different factors, compared to what is known at other
nar debitage waste products; the concentrations differ only regional and extra-regional Bandkeramik settlements, al-
in the form of their rejection and their exclusive composi- low us to define a new production economy. In the context
tion of laminar debitage waste products (absence of tools, of studies that we are currently conducting to document
burned pieces, flake debitage, etc.). This spatial distribu- this phenomenon, we address here the problem of quanti-
tion indicates that there is no specific location for blade fying the objects produced and consumed on the scale of
production activities and that this production is distributed the spatially distinct production units that correspond to
among all of the households. concentrations, houses and villages.
- The nearly exclusive procurement of raw materials from
sources in direct proximity to the site. 3. Method to quantify laminar produc-
- A laminar production with a single intention oriented to- tion: preliminary considerations
wards blades of average to fine regularity, with parallel
edges and a rectilinear profile. Blade length is standard- Any attempt to quantify the products yielded by a blade
ized, as is clearly shown by the core dimensions. The in- core is a particularly perilous task. We must first empha-
tended range is 80120 mm in length, 1530 mm in width size the major difficulties associated with this calculation:
and 48 mm in thickness. - Flint nodules, Campanian flint in this case, have highly
- A rational debitage management. The extraction of blades variable forms and dimensions; the volumes of the blocks
is organized according to two principal procedures. The to be knapped thus cannot be considered as equivalent.
most frequent consists of removing blades from the two Therefore, calculations of the number of blades produced
edges of the debitage surface, in order to extract a central by each block must be made case by case, based on ex-
blade, resulting in a 2-1-2 rhythm. This rhythm was ob- haustive refits.
served on over 50% of the cores (Allard 2005). The second - The quality and homogeneity of the raw materials is
33
highly significant, since imperfections or irregularities b) Among the well preserved concentrations, extensive re-
in the flint directly influence the core preparation and refits could be made in structure 130. The proportion of refit
duction, and consequently the quantity and quality of the pieces over 2 cm is 29% (1243 out of 4272), represent-
blades produced. This problem increases if several distinct ing 75% of the weight of the assemblage (74,600 kg for
flint types are exploited. 98,513 kg total weight).
- The nature of the production is crucial: indeed, if the objec- c) Only one type of flint was included in the concentra-
tives do not tend toward standardized products, at least in terms tions: the fine-grained, gray Campanian flint of Hesbaye.
of dimensions, their quantification becomes too speculative. However, this flint is not always of good quality: its nu-
- Finally, the uniform application of a debitage technique merous imperfections sometimes presented difficulties, or
is also essential. According to the results of experiments even interrupted the core reduction process.
conducted by Jacques Pelegrin (personal communica-
tion), pressure debitage produces the most standardized 50 Core length
products, though indirect percussion also produces fairly 45 43,68
standardized blanks. 40
35
30 26,92
4. The quantification of debitage prod- 25 21,15
ucts at the scale of spatial units 20
15
10
The problem of quantifying debitage products in concen- 5
5,77
1,37
trations, houses, and entire villages can be addressed at
0,55 0,27
0
Verlaine based on the analyses that we propose here. The
65-80 80-95 95-110 110-125 125-140 140-155 155-170
mm
problem is clearly not to quantify the exact number of
blades produced, such an objective being inaccessible in Fig. 2: Verlaine Petit Paradis. Dimensions of the cores of feature 1.
our current state of research, but rather to formulate ac-
ceptable estimates.
d) The dimensions of the blade products, and especially
4.1. Characterization of waste products in concentra- the cores, are highly standardized. Among the 439 cores in
tions structure 1, 85% were abandoned with a blade surface of 8
to 12 cm and 44% at 9.5 to 11 cm (Fig. 2).
The data from concentrations at Verlaine present charac- e) The debitage of blade products was made by indirect
teristics that justify their quantitative exploitation: percussion (Allard op.cit.). The macroscopic stigmata are
a) These concentration waste products constitute closed particularly clear in this assemblage. The principal charac-
productive assemblages, which have the following char- teristics we observe are the absence, or only slight prepa-
acteristics: ration, of the striking platform, often leaving a protruding
1- Homogeneity of rejected objects and their burial (at edge, smooth and sometimes concave butts, a striking an-
least for the smallest pieces): the concentration of lithic gle of 80-90 and the frequent appearance of a half-circle
remains is homogeneous, regardless of the archaeological behind the butt.
structure in which they are contained (only flint, no mix f) The refits and debitage characteristics show that these
with other materials); were good knappers, notably in their ability and flexibility
2- Spatial and temporal homogeneity of the production: to adapt the optimal debitage of standardized blades to the
the abandoned concentration corresponds to a strictly co- initial morphology of the flint nodules, the low number of
herent production, both temporally and spatially; rejuvenation flakes per block, and especially in their re-
3- High standardization of the blanks produced: the sponses to imperfections in the flint (Allard 2005, 2007).
blades are produced in series, and most correspond to re- We thus observed efficient and regularized block reduc-
current standards (see d) below). tions performed by experienced knappers.
34
4.2. The methods of production calculation applied to This debitage served as first unit of measure to estimate
the concentrations the number of products obtained from the cores of con-
centration 130.
Our reflections were based on the refits of the concentra-
tion of structure 130. There are few complete refits, but the This estimation is complicated by two problems:
partial ones, notably those associated with repairs to core The experimental blades are slightly narrower and thin-
striking platforms, are systematic, except of course for the ner than the retouched products of the Verlaine structures
blocks with natural striking platforms. (Fig. 3). In fact, their dimensions correspond to those of
the laminar products found in the concentrations, i.e. to the
The method to assess the total number of blades follows totality of rejected blades and broken standardized blades.
several steps. In this paper, we present only the first at- It is thus probable that the number of experimental blades
tempts, which are based on a few cases. In optimal situ- is greater than that of the Bandkeramik cores.
ations, we applied three distinctive approaches, but more The quality of the experimental debitage shows that the
generally two. At this stage of analysis, only one has been standardized blades were removed one after the other with
systematically developed. These calculations are based few maintenance products or small correction bladelets.
on the laminar debitage volume observed through refits This is due to the excellent initial preparation of the block.
of core tablets, and the opening of striking platforms on At Verlaine, the preparations are generally less thorough,
cores (which shows in negative the maximum volume and the maintenance products are more numerous. This
exploited). From a methodological perspective, the refits probably means that the number of experimental blades is
that correspond to repairs and the opening of striking plat- a little too high.
forms are sufficiently pertinent to enable us to attempt an
estimation of the number of blades produced. In addition,
they are advantageous because they are relatively easy
to find in all the concentrations, even the most eroded.
However, the negative volume visible between the open-
ing of the striking platform and the core or core tablet cor-
responds to the total volume exploited, including the small
preparation flakes, platelets, and other products. It thus
corresponds not only to the standardized blades, but to all
laminar debitage products.
Experimental debitage (method 1)
The experimental debitage was carried out by Jacques Pe-

legrin (CNRS-UMR 7055). He worked with a block of fine-
grained Hesbaye flint, around 15 cm long, collected during
the excavation of test pits in the dry valley that runs through
the Bandkeramik site of Verlaine. The laminar volume corre-
sponds to that estimated for the archaeological pieces (prod-
ucts 8-12 cm long). The debitage began with the creation of
a frontal crest, and developed over a surface of an average
of five blades. A core tablet was removed after a first pro-
duction phase of around 30 laminar products. The five last
blades were short and flat. The products were measured, and
compared with the blades in concentrations and other struc-
Fig. 3: Verlaine Petit Paradis. Dimensions of the experimental blades
tures, as well as with the tools from Verlaine (Allard 2005). and of the archaeological blades.
35
Theoretical reconstruction based on blade dimensions The most probable theoretical scheme yields a total of 16
and the reduction of volumes (method 2) blades, but that includes many small ones. Confrontation with
the experimental debitage yields a total of 16-17 blades.
The second method corresponds to a theoretical ideal With these parameters, we thus obtain a difference of 2
scheme of the exploitation of laminar volumes (for exam- to 3 blades relative to core 18, or 14% in this particular
ple Astruc et al. to be published). The known parameters case (6% between the experimental and the theoretical
are the rhythm of the final debitage of the core (reading scheme). This difference is clearly explained by the knap-
of a diacritic diagram of the laminar surface), the aver- pers management of the volume, beginning the debitage
age width and thickness of the products, the section of the with a thick blade prepared by bladelet removals, and the
intended products and the negative of the laminar volume. preparation of the left side by the removal of short blade-
With these parameters, it is possible to redraw an exploita- let-blade products (Fig. 4A).
tion scheme of blades for each core with average laminar
sections. Of course, this scheme applies only to a theo- Case 2: In this case, we tested the theoretical and experi-
retical ideal, because it represents only perfect products. mental schemes (Fig. 4B). We obtained 23 blades for the
It does not account for imperfections in the flint that can first and 25 for the second (8.7% difference).
modify the debitage process, knapping accidents or small
laminar products corresponding to the maintenance of the Case 3: Finally, a last attempt was made with core 65 (Fig.
core, or the regularization of its sides, for example. 4C), on which we were able to refit the opening blade and
This method is nonetheless pertinent in the context of a a cortical blade. It allowed us to rather precisely refine the
standardized core reduction made by indirect percussion, theoretical scheme to 8-9 blades as opposed to 9 for the
such as at Verlaine. experimental debitage (12,5% difference).
Refits (method 3) 5. Results
In addition to pieces associated with core preparation and In this paper, we propose only a first estimation of the
the opening and repairing of striking platforms, whose reproduction, because all three methods were not systemati-
fitting allows us to reconstruct the laminar volumes, nu- cally applied to the assemblage of cores of concentration
merous blade fragments (620) are also present in the con- 130. The only exhaustive estimation conducted was that
centrations. In two cases, most of the blades could be refit, of experimental debitage. This was possible for 58 blocks
thus allowing an exact restitution of the debitage (both cas- out of 66, because some had no production: they had been
es are blades broken in the middle during debitage). These abandoned before the extraction of standardized blades,
are the most pertinent refits, as they give a true vision of while others were impossible to estimate because the strik-
the range of blades produced and the small adjustments ing platform was natural and never repaired. Finally, a few
made on the sides of the core or its laminar surface. cores collected by Neolithic people themselves are miss-
ing from the total assemblage.
4.3. Application
The 58 blocks that furnished laminar blanks correspond
These different approaches were tested on a few cases in to 65 distinct cores. Based on the method of experimental
order to compare estimations of the number of blades pro- debitage, an average of 21 to 23 blades were produced per
duced. core. The estimation of the number of blades produced thus
Case 1: We were able to apply a triple approach to core 18, oscillates between 1365 (65x21) and 1495 (65x23). This
thanks to the 5 blades that could be refit onto it. The ar- is a minimum interval, since for at least 5 cores we have
chaeological debitage shows the extraction of 14 blades. The only the second phase of exploitation, which is posterior
sequence corresponds perfectly to that most often recorded to a core tablet. By reference to the cases described above,
on the cores: the extraction of lateral blades to prepare the the difference between the experimental and theoretical
central blade with three sections, in a 2-1-2 rhythm. scheme is roughly 1 (cases 1 and 3) to 2 (case 2) blades
36
core n 18
core n 18
5
3 in black refited blades
2
1
core n65
core n 2
B C
Fig. 4: Verlaine Petit Paradis. The production calculation applied to 3 cores of feature 130.
Restitution of laminar debitage with the refited blades (method 3) and restitution with theoretical scheme
of the exploitation (method 2) n 18, 2, 65.
per core. By extrapolation, this gives from 19 (21-2) to 21 was attempted by counting the whole and almost whole
(23-2) blades per core for the lowest range, and from 20 to blades and the distal parts (the most numerous), which cor-
a maximum of 22 specimens, for a total number of blades responds to 327 products.
within the range of 1235 (19 x 65) /1365 (21 x 65) to 1300 Therefore, in concentration 130 at Verlaine, for the pro-
(20 x 65) /1430 (22 x 65). duction of the 58 blocks accounted for, we suggest that
between 900 and 1200 blades were collected by the Neo-
Finally, by comparing with the only archaeological refer- lithic inhabitants.
ence in this example, the difference would be from 2 to
3 blades, which would give 18 to 20 blades per core, and The relationship between the quantity produced
thus a total number of 1170 to 1300 specimens. and the quantity used in situ
These three estimations are in fact pretty close, and a range The problem of the quantification of products consumed
between 1200 and 1500 blades seems correct when we con- in spatial units of the village can be only partially treat-
sider the three possibilities, or between 18 and 23 blades ed here, as the study of artefacts from the site is still in
per core. This range is approximate but gives an idea of progress, and the methods of calculation are currently be-
what can be done by confronting methods and data. ing developed.
The concentration contained 620 whole or fragmented Nonetheless, the stratigraphic position of the concentra-
blades. A calculation of the minimum number of items tion of structure 130 offers a good opportunity to study
37
this production within a domestic unit. The assemblage ultimately the site. We will also analyze the qualitative
was located at the bottom of the structure, with no relation variability of the blades produced and used. Finally, we
to the upper detritic level of the pit lateral to the habitation must consider the entire functional spectrum, to reveal the
unit. We can legitimately suggest the hypothesis that the relative proportion of the exploitation of the sub-products
concentration resulted from the fabrication of a stock of of laminar debitage. These parameters enable us to bet-
blades for the inhabitants of this house. We thus analyzed ter define the dichotomy between an overabundant pro-
the objects in the domestic pit made from the fine-grained duction whose objective is not a functional intra-muros
Hesbaye flint and compared them with those of the assem- exploitation, but export; simultaneously, a weak in situ
blage. consumption of products corresponding to a reduction se-
This second assemblage differs in that it contains tools (in- quence identical to the preceding one, applied to the same
cluding cores reused as hammerstones), used flakes and raw material. Integrating these different aspects will allow
burned objects, artefacts already observed in other similar us to better conceptualise the complexity of the objectives
structures in this site. which informed the management of the production of tool
The artefact assemblage of the detritic level also includes blanks in the Bandkeramik of Hesbaye.
the ensemble of laminar debitage waste products, in pro-
portions comparable to that of the concentration. The only
real difference lies in the number of bladelets and lami- Pierre Allard
nar flakes, but this could be related to the archaeological CNRS-UMR 7055 Prhistoire et Technologie
technique (the concentration was removed in one block 21 alle de lUniversit
and sifted). Elements of all stages of laminar debitage F-92023 Nanterre Cedex
are present, from unworked blocks to cores. There are 18 e-mail: pierre.allard@mae.u-paris10.fr
cores, including three reused as hammerstones and three
fragments. Retaining the estimation proposed based on the Laurence Burnez-Lanotte
concentration, we obtain a range from 378 (18 x 21) to 483 University of Namur FUNDP
(23 x 21) blades for 21 cores, or, if we consider only the Laboratoire de Protohistoire europenne
pieces not reused as hammerstones, from 144 (18 x 8) to 61 rue de Bruxelles
184 (23 x 8) blades. B-5000 Namur Belgique
There are 154 unretouched and retouched blades in the de- Associe UMR 7041 CNRS
tritic level. This number is thus equivalent, or even inferior e-mail: laurence.burnez@fundp.ac.be
to the number estimated based on the debitage waste of
this assemblage.
References
Conclusion
Allard, P. 2005. Lindustrie lithique des populations ruba-
The test that we have proposed is based on the integra- nes du nord-est de la France et de la Belgique. Interna-
tion of three methods applied to the artefacts of a debit- tionale Archologie, 86. Rahden/Westphal: Verlag Marie
age concentration and of detritic waste found in the same Leidorf GmbH.
structure. The estimation of the average minimum number
of blades produced per core is 18 to 23, resulting in a mini- Allard, P. 2007. Surplus production of flint blades in the
mum of 1200 to 1500 blades for the entire pit. In contrast, early Neolithic of Western Europe: new evidence from
the unretouched or retouched blanks of this assemblage, Belgium. European Journal of Archaeology 8, 2005, 205-
or blades used intra-muros, represent approximately 40% 223.
of the blades in the detritic levels and 10% to 20% relative
to the minimum number of blades produced based on the Allard, P. and Burnez-Lanotte, L. 2007. Surplus produc-
debitage waste of the entire pit. This large deficit compels tion in the Belgian Linearbandkeramik: blade debitage at
us to further quantitative analyses in the entire house and Hardumont Petit Paradis (Verlaine, Hesbaye, Belgium).
38
In Stone Age-Mining Age, VIII International Flint Sympo-

sium. Bochum, September 1999, Der Anschnitt beiheft 19.
Bochum, 37-54.
Astruc, L., Gartuze, B., Pelegrin, J. and Akkermans, P. in

press. From Production to use: a parcel of obsidian blade-
lets at Sabi Abyad II. In L. Astruc, D. Binder and F. Briois
(eds.), La diversit des systmes techniques des communau-
ts du Nolithique pr-cramique: vers la caractrisation
des comportements sociaux, 5me colloque international
sur les industries lithiques du Nolithique pr-cramique.
Antibes: ditions APDCA 2007.
Binder, D. and Gassin, B. 1988. Le dbitage laminaire

chassen aprs chauffe: technologie et traces dutilisation.
In S. Beyries(ed.), Industries lithiques, tracologie et tech-
nologie. B. A. R., Int. Ser., 411, Oxford : BAR, 93-125.
Burnez-Lanotte, L. and Allard, P. 1998. Mode de produc-

tion laminaire dans le Ruban de Hesbaye au Petit Paradis
Hardumont (Verlaine, Lg). In Organisation nolithique
de lespace en Europe du nord-ouest, XXIIIme colloque
interrgional sur le Nolithique, 24-26 octobre 1997
Bruxelles, Anthropologie et Prhistoire, 109, 15-26.
Burnez-Lanotte, L. and Allard, P. 2003. Blade debitage in

the Belgium Linearbandkeramik: the production at Har-
dumont Petit Paradis (Verlaine). In L. Burnez-Lanotte
(ed.) Production and Management of Lithic Materials
in the European Linearbandkeramik, UISPP, Septembre
2001, Lige, Belgique, B.A.R. Int. Ser. 1200, Oxford:
BAR, 59-64.
Acknowledgements
We are very grateful to the Fyssen Foundation for the in-

terest shown in our lithic research and for funding Pierre
Allard a post-doctoral year at Namur. We would also like
to thank Jacques Pelegrin (CNRS-UMR 7055) for his
comments and his help about the theoretical idea scheme
of the exploitation of laminar volumes.
39
40
Hlne Collet, Anne Hauzeur, Jacek Lech: The prehistoric flint mining complex at Spiennes on the occasion of its discovery 140 years ago
The prehistoric flint mining complex

at Spiennes (Belgium)
on the occasion of its discovery 140 years ago
Hlne Collet, Anne Hauzeur, Jacek Lech
Abstract: 140 years ago, the Neolithic flint mining complex at Spiennes, Hainaut province, became known to Europe
and the world. That discovery spurred the dynamic development of research into prehistoric flint mining in European and
world archaeology. This anniversary is an occasion to sum up what we know today about this mining complex which, in
the year 2000, was placed on the UNESCO world heritage list.
Excavations of the Spiennes flint mine sites have been a long and rich tradition producing important results for European
prehistory. These results are presented in the first part of the paper. In the second part, the authors sum up the current state
of knowledge about the mining fields at Spiennes, about the various methods of flint exploitation, production trends and
the chronology of mining activities. Particular attention has been paid to studies of old chipped inventories and to the most
significant results of recent research led by H. Collet.
Keywords: Spiennes, Neolithic, prehistoric flint mining, methods of chipping floor research, flint cores and blades, flint
axe, Michelsberg culture, history of archaeology.
Introduction
Spiennes is one of the most important names on the map 19th century and the first decades of the 20th, the Spiennes
of prehistoric flint mining. Today, the village lies on the micro-region became famous as a place where many im-
outskirts of Mons, the capital of the Hainaut province in the portant discoveries connected with prehistoric flint mining
region of Wallonia, Belgium (Fig. 1). For ages, the place had been made, influencing developments in the archaeolo-
has been known for the vast number of flints lying on the gy of prehistoric flint mining in Europe and North America.
surface of the fields. Towards the end of the first half of When discussing the places where flint tools were made,
the 19th century, collectors of antiquities became more and John Evans (1872: 29-32) compared Spiennes to Grimes
more interested in these flint pieces. While Spiennes fame Graves and Cissbury in England, and to Grand-Pressigny
was not equal to that of Grand-Pressigny in France, it was in continental Europe. His views were in accordance with
visited by Belgian and foreign antiquarians who collected those of Albert Toilliez, who first studied Spiennes. Today,
ancient flint artefacts. the prehistoric mining complex at Spiennes is considered
140 years ago, in 1867, a spectacular archaeological discov- one of the most important features of prehistoric flint min-
ery was made at Spiennes. A railway cutting revealed the ing in the world, next to Grimes Graves and Cissbury in
deep shafts of a Neolithic flint mine and its underground England, Rijckholt-St Geertruid in the Dutch Limburg, and
galleries. Spiennes became known to the world. The 1867 the Krzemionki Opatowskie in Poland. It has been on the
find was followed by others, so that in the second half of the UNESCO World Heritage list since November 30, 2000.
41
Fig. 1: Location of the prehistoric mining complex at Spiennes, Hainaut province (Belgium) and other important archaeological sites.
1. Spiennes flint mine. 2. Thieusies. a - Middle Neolithic settlements and flint concentrations; b - Middle Neolithic enclosure sites; c - flint mines; d -
loess area; e - Spiennes flint distribution; f - supposed Spiennes flint distribution. After Vanmontfort et al. (in print).
1. Context of the mining field at Spi- crest with asymmetric slopes extending from east to west,
ennes which culminates at a height of 93 m. The cuesta is cut
by the river La Trouille and the river La Wampe, tributar-
1.1. Geography and geology ies on the left side of the river Haine. The substratum of
this cuesta consists of Upper Cretaceous chalk covered by
The mining field at Spiennes is located in the district of Tertiary glauconitic sands and Wechselian loess. From top
Mons, about 5 km south-east of the city of Mons, in the to bottom, the chalk formations are Craie de Spiennes,
Belgian Province of Hainaut. The centre is situated 123,26 Craie de Nouvelles, and Craie dObourg. The Spiennes
N and 123,19 E by the Belgian Lambert coordinate system formation, of which the basal 10-12 m date from the
and 50 25 11,01N and 3 59 29,99 E by the geographi- upper part of the Upper Campanian (Robaszynski, Chris-
cal coordinate system WGS84. tensen 1989), includes many thick black flint seams. The
Neolithic miners exploited several of these.
From a geographical and geological point of view (Fig.
2), the mining field belongs to the Mons basin, which is 1.2. The mining field
a west-east oriented sink where the river Haine flows, a
tributary of the river Scheldt. To the north, it is surrounded The site is located in fields south of the village of Spiennes
by the low plateaus of Hainaut, that of Anderlues to the (Hainaut, Belgium) occupying the slopes of La Trouille
east and the Haut-Pays to the south. The mining site is lo- valley, parts of the two plateaus surrounding it and the
cated on the northern flank of the Harmignies cuesta, a slope toward the river La Wampe (Fig. 3). Flint mines and
42
Fig. 2: Geological map of the Spiennes flint mine. Design and drawing by O. Collette. a extent of flint mine based on surface mined
flint distribution plotted by F. Gosselin.
Fig. 3: The prehistoric mining complex at Spiennes. Map of the site with the most important places excavated so far.
a - 1867 railway trench; b - shafts; c - workshops; d - geological anomalies; 1 - 1925 A. de Lo and E. Rahir excavation; 2 - approximate
location of a gallery found by L. De Pauw and E. Van Overloop in 1889-1890; 3 - 1912-1914 A. de Lo excavation of deep shafts 1 and
2; 4 - approximate location of 1930 J. Breuer excavation; 5 - 1953 J. Verheyleweghen excavation in the slope of the La Trouille river
valley; 6 - SRPH excavation since 1953; 7 - 1965 F. Hubert excavation; 8 - interruption of the Michelsberg enclosure; 9 - shafts anterior
to the Michelsberg enclosure; 10 - 1975 F. Hubert excavation of shafts and mining pits near the railway trench; 11 - 1990 F. Hubert and
M. Soumoy excavation of shallow shafts; 12 - 1997-2004 H. Collet excavation at Petit-Spiennes; 13 - 2005 H. Collet excavation of
chipping floors on plot 51c; 14 - 2005 H. Collet excavation in the slope of the La Trouille river valley; 15 - shafts discovered by the sci-
entific society Recherches et prospections archologiques en Wallonie during the laying of a gas pipeline; 16 - 1962 J. Verheyleweghen
excavation of a shaft; 17 - evidence of shafts discovered in 2004 by H. Collet on plot 169a east of Harmignies Street.
43
workshops cover about 100 hectares (Hubert 1980: 124). Mont de Prle. These extracting features were mentioned
The mining site can be divided into three different mining in local newspapers at the time, but their date was un-
areas. The largest one, called Camp--Cayaux (about 65 known.
ha), is located on the plateau east of the river La Trouille. As early as 1851, Albert Toilliez suspected the probable ex-
The second, known as Petit-Spiennes (about 14 ha), is situ- istence of an axe workshop at Spiennes. His cousin Dsir
ated on the plateau west of La Trouille and extends paral- Toilliez reported that Toilliez had collected more than 60
lel to the river. A third one, named Versant de La Wampe, flint axes, and roughouts, in the district of Harmignies and
discovered later, is located on the plateau of Petit-Spiennes in the district of Spiennes, where these artefacts may have
in the part surrounding the river La Wampe. This is also been manufactured (Toilliez 1851).
where an enclosure was built during the Neolithic period. Shortly after, a date later than Palaeolithic was proposed
for these workshops, despite the fact that some researchers
2. Historical background at that time held to the older chronology. This conviction
was not based on stratigraphic evidence but only on similar
2.1. The first discovery by Albert Toilliez morphology between hand axes and axes roughouts from
Spiennes (De Koninck 1860).
Albert Toilliez (1816-1865), a mining engineer in Mons, After A. Toilliezs death in 1865, his collection seems to
was the first important researcher at the Spiennes archaeo- have been bought by the famous British antiquarian, field
logical site (Fig. 4). archaeologist and numismatist Sir John Evans (Hamal-
Before this 1850s discovery, some shafts and galleries, Nandrin, Servais 1925: 75), and then sent to Oxford.
most probably Neolithic, had been observed for the first
time in the summer of 1842, near the southern and western 2.2. The railway trench in 1867
rim of the plateau of Camp--Cayaux, at a place called
In 1867, the building of the railway from Mons to Chimay
required the digging of a deep trench to cross the plateau
of Petit-Spiennes (Fig. 5). Thus were discovered several
spectacular cross sections of flint mine shafts in the wall
of the trench.
In 1868, the controversy over the chronological position

of the Spiennes artefacts came to an end with the pub-
lication of the report sur les dcouvertes gologiques et
archologiques faites Spiennes (Briart et al. 1868). The
authors claimed that there had been two periods of human
occupation at Spiennes, with the extensive workshop and
the digging of shafts occurring during the later period. The
numerous chipped flakes found in the surroundings of Spi-
ennes had been made in situ with materials extracted in the
vicinity.
The dating of the shafts was determined on the basis of

polished axes, potsherds tempered with flint, and bones
of recent fauna, including those of the domesticated dog.
This referred to the workshops of Grand-Pressigny, and
to the fact that a dagger was discovered at the surface at
Fig. 4: Mons, Hainaut province. Grave of Albert Toilliez (1816-1865)
the first important researcher of the Spiennes flint mining complex. Pho- Spiennes. At that time, this artefact was thought to have
tograph by A. Hauzeur.
44
Fig. 5: Spiennes, Hainaut province. Petit-Spiennes. Cross section of the 1867 railway trench. After A. Briart, F.-L. Cornet and A. Houzeau de Lehaie (1868).
Fig. 6: Spiennes, Hainaut province. Camp--Cayaux. Cross section of

the shaft excavated in 1887 by A. de Lo and E. de Munck. After A. de
Lo and E. de Munck (1891). A: Humus with flint flakes; B: Flint flakes,
cores, blades, axe roughouts, small fragments of chalk, and chalk pow-
der; C: Spiennes chalk; C: Large chalk block with two lateral grooves of
a rope; C: Flint seams; D: Blocks of chalk, small fragments and powder,
flint nodules, large flakes, and several worked pieces.
been made of Grand-Pressigny flint1. For the authors, this

demonstrated exchanges between both workshops (Briart
et al. 1868: 379).
The results of the excavations were presented by Alphonse

Briart (1825-1898), Franois-Lopold Cornet (1834-1887),
and Auguste Houzeau de Lehaie (1832-1922) at the Interna-
tional Congress of Anthropology and prehistoric Archaeol-
ogy held in Brussels in 1872. Publication in the proceedings
gave the study and the site of Spiennes some celebrity.
1. Today this piece is surmised to be the dagger from the Houzeau de

Lehaies collection. According to M. Delcourt-Vlaeminck, it was not
made of Grand-Pressigny flint (pers. comm.).
45
In spite of its importance, the discovery was not followed excavations headed by Jacques Breuer around the 1950s.
up by a serious programme of research and excavation. An area of about 250 m was investigated. A very detailed
The only subsequent investigations were the excavation of and exhaustive study of these shafts was conducted by F.
a mine shaft - not the galleries - undertaken in 1887 by Hubert in 1978, with a detailed underground map, a fine
Baron Alfred de Lo (1858-1947) and Emile de Munck stratigraphy of the geological background, and a descrip-
(1861-1944). Those excavations are situated in the heart tion of the picks (Hubert 1978). New excavations were
of Camp--Cayaux to the south west of the present Centre conducted in 1982 by F. Hubert (1983) who reported on
de Recherches archologiques. The first, schematic, cross his findings but did not offer a detailed study of the ar-
section of the shaft excavated shows that the flint seam ex- chaeological material.
ploited appears to be the sixth (Fig. 6). Moreover, excava-
tions explored some upper parts of shafts, workshops and 2.4. Field research
other pits. The results were presented at the International
Congress of Prehistoric Anthropology and Archaeology Baron Alfred de Lo and Edmond Rahir also excavated
held in Paris in 1889 (Lo, Munck 1891). in 1925 some 34 sunken-floor huts as they called them
(Lo, Rahir 1929); actually, these should be considered
This was also the period of the first technological studies heads of shafts. Similar surface explorations of the site
of the material, such as a suggested method of blade pro- were conducted in 1924 by Jean Hamal-Nandrin, professor
duction by indirect percussion. These observations were at the State University of Lige, and Jean Servais, curator
supported by the discovery of bone punches (Cels, De of the Archaeological Museum of Lige (Hamal-Nandrin,
Pauw 1885-1886). Louis De Pauw and Eugne Van Over- Servais 1925). Others were made by Charles Stevens in
loop (1889-1890) noted, for the first time, a gallery dug the beginning of the 1920s for Aim Rutot, curator of the
directly into the slope of the valley, below the western part Museum of Natural History in Brussels. Charles Stevens,
of Camp--Cayaux. a local quarry worker, was active over a very long period
(more or less between 1911 and 1953), and excavated on
2.3. The excavations of Baron Alfred de Lo his own or on request. He himself chose the places to ex-
cavate. As he left no record of where exactly he had been
The excavation of two shafts at Camp--Cayaux by the digging, nor any drawings, we have hardly any informa-
National Excavation Service, based at the Muses du Cin- tion about the features he excavated, neither their mor-
quantenaire, began in July 1912 thanks to a grant from phology nor the nature of the fillings. His main purpose
Count Louis Cavens. These truly ambitious excavations was to enlarge the collections of museums and collectors,
were conducted by Baron Alfred de Lo and continued in with no concern for such details as contexts. His research
1913 and 1914 (Lo 1925). Two shafts, 16 m deep, were ex- has been published only partially, if at all.
cavated in 1912 (Lo 1925; Lo, Rahir 1929). Parts of the
galleries were explored in 1913, as well as some workshops In 1930, Jacques Breuer (1892-1971) applied new and
in the vicinity of the shafts. More than 1,500 picks were improved techniques of excavation. He dug real archaeo-
collected during these campaigns, as well as several ham- logical test trenches to systematically determine the kind
mers in Landenian or Devonian sandstone. Thin potsherds of structures. Breuer identified features such as temporary
tempered with very fine flint chips were collected. They open air mining zones. He drew attention to the scarcity of
are part of the chalice-shaped pots, attributed by Franois artefacts, especially pottery and domestic remains (Breuer
Hubert to the Michelsberg Culture (Hubert 1978). 1930).
Pictures of the galleries and simplified cross sections of the 2.5. Jean Verheyleweghens excavation and interpreta-
two shafts, with the drawing of some layers, were published. tion
Unfortunately, only general information was given about
the archaeological material (for example, Lo 1925). This tradition of surface test trenches was also practised
More galleries were later explored by Maurice Lefort, in by Jean Verheyleweghen (1910-1965) who excavated
46
nearly 250 features between 1945 and 1958. These were naut. The localisation of the excavation site was chosen
interpreted as knapping and chipping places, funnels of by Andr Adam (1919-2005) who had noticed concentra-
shafts, some occasional pits of domestic waste, and one tions of flint flakes, waste, and fragments of chalk on the
secondary burial. The structures were widespread over the ground surface. A first shaft and the upper part of a second
whole surface of Camp--Cayaux, and along the slope of one were excavated at this site, until 1960. After a pause,
the right bank of river La Trouille (Verheyleweghen 1963). research started again in 1979, at the same site exploring
other shafts and underground galleries. Work here is still in
progress, with 95 m so far excavated.
Marcel Lefrancq (1916-1974) with the help of Paul-Henri
Moisin (1927-2004), both members of the Society, led the
excavations, supported by professionals archaeologists.
Their work and publications were of excellent scientific
quality, even if we deplore the absence of drawings of
cross sections.
Between 1988 and 1992, excavations were carried out by
Jean-Pierre Joris with the help of Franoise Gosselin, who
published the first monograph of the Petit-Spiennes area,
including all the results since the beginning of excavations
in 1953 until the date of publication (Gosselin 1986).
Fig. 7: Spiennes, Hainaut province. Camp--Cayaux. The slope of the La
Trouille river valley. Cross section of flint exploitation remains found in 2.7. Franois Huberts research and excavations
1953 by J. Verheylewhegen: a - disturbed sediment; b - humus; c - yellow
sediment mixed with chalk in blocks and powder; d - natural chalk rock
and flint seams; e, f, g - workshops according to J. Verheyleweghen.
The next important step in the research at Spiennes was
After J. Verheyleweghen (1953).
done by Franois Hubert, assistant in the Service national
des Fouilles at that time. He undertook excavations both
These important field campaigns allowed him to propose in the Camp--Cayaux area and on the Petit-Spiennes pla-
an outline of the development of the mining field at Camp- teau.
-Cayaux. This schematic evolution is divided into four In 1965, Franois Hubert excavated four workshops as well
periods. The first three developed from a simple excavation as the upper part of three shafts at Camp--Cayaux. His
along the slope of the valley to the complex mining works finds included some Michelsberg pottery (Hubert 1969). A
of the deepest shafts. The fourth phase corresponds to the field survey allowed him to determine more precisely the
decline of the mining activities. The outline was taken as a extent of the mining field. In 1975, he conducted a rescue
model to interpret the flint mine site of Rijckholt St-Geer- excavation initiated by the construction of a gas pipeline
truid (NL), and Cissbury (GB). Despite its importance, this crossing the railway at Petit-Spiennes (Hubert 1976a). All
work has never yet been critically discussed, mostly be- these excavations were highly professional and provided
cause of the lack of later large-scale excavations and accu- interesting data about material, stratigraphy and methods
rate chronology. Verheyleweghens results and ideas were of exploitation.
fully accepted by Siegfried J. De Laet, and published in his
book La Belgique davant les Romains (1982). Since January 1954, Andr Adam, one of the founders of
the Socit de recherche prhistorique en Hainaut, had
2.6. The beginnings at Petit-Spiennes suspected that a settlement existed on the plateau of Petit-
Spiennes south of the railway. Indeed a Michelsberg enclo-
The first excavations of the Petit-Spiennes plateau began sure was in fact discovered right at that spot (Adam 1959).
in 1953. They were organised by several members of the The same year, Irwin Scollar (1955) noticed the presence
society Les Amis du Muse de Mons which later became of crop-marked fragments of two concentric circles when
known as the Socit de recherche prhistorique en Hai- he examined aerial views. Between 1966 and 1979, F. Hu-
47
bert carried out annual excavations in the ditches, by mak- by the volunteers of the SRPH since 1953 continues to be
ing trenches at regular intervals. Only the first years of his explored. A new mine connected with the previous ones
investigations have been published (Hubert 1971), as well has been excavated and the filling of two shafts is currently
as the ceramic vessels (Vanmontfort 2004). being explored (Desterbecq 2004; Desterbecq, Joris 2006;
Robert 2003).
During the annual excavation of the ditches of the forti-
fied camp in 1970 and 1974-1975, flint extraction features 3.1. Rescue excavations at Petit-Spiennes and research
(shafts and gallery roof), earlier than the ditches, were dis- results
covered (Hubert 1971: 50-51; Hubert 1976b: 12-15). The
inner area of the fortified camp is still almost completely In the case of systematic excavations, it was decided to
unexplored. Only one trench 2 m wide and 108 m long was record the data as fully and carefully as possible. For this
dug inside the camp. reason, the choice was made to draw very detailed cross sec-
tions of the filling of the shafts. In the case of rescue excava-
3. The latest decades of research at tions, a strategy of recording of arbitrary 20 cm levels was
Petit-Spiennes and Camp--Cayaux adopted; later however, this was replaced by recording the
real observed stratigraphic layers. Detailed stratigraphy of
Several kinds of explorations took place at Camp--Cayaux the extraction features made it possible to better establish
during the 1990s, giving us a better idea of the extension the different stages of filling and to try to calculate the speed
of the mining area. Preventive excavations around the of the filling process. To confirm or to help understand this
Centre de recherches archologiques yielded five shallow question of deposit speed, stratigraphic data were crossed
shafts (Hubert, Soumoy 1993; Bonenfant, 1993). Rescue with the paleoenvironmental data: palynological, archaeo-
excavations, prior to the laying of a gas pipeline, allowed zoological (primarily microfauna and snail remains), and
for the detection of 20 more shafts located in the northern anthracological (Collet, Van Neer 2002; Defgne, Collet
part of the mining site, and contributed to the extension of 2003).
the known mining area in this direction. A small Neolithic
workshop was excavated in the south-western part of the 3.2. The discovery of a chipping floor at Camp--
mining site, in the valley of La Trouille (Fechner et al. Cayaux
1993).
For some years now, rescue and survey excavations have
Preventive excavations were also undertaken at the min- been conducted in the surrounding area, in the vicinity of
ing site of Petit-Spiennes, because there was a project to the village of Spiennes, prior to the building of a fence,
develop the archaeological site (Collet et al. 1997). Luck- house or renovation of the garden of the Centre de recher-
ily the project was put aside, and the excavations turned into ches archologiques.
planned ones, led by Hlne Collet. Since there is no emer- In early 2005, the archaeological team undertook a res-
gency in this area, and rescue work is needed much more cue excavation, prior to the building of a house, on lot 51c
urgently elsewhere, these excavations are still continuing. which lies just outside the part of the site under Cultural
Heritage protection, but inside the archaeological site of
Several features were identified, among which some mine Camp--Cayaux. During this investigation, 17 extraction
shafts should be noted. A deep shaft was fully excavated, features were discovered over an area of about 150 m, as
as well as half of another; there was also what was inter- well as remains of in situ workshops.
preted as an aborted shaft, as well as shallow pits. So far
only provisional papers have been published regarding this 4. Spiennes and research concerning
research (Collet 2000; 2003b; Collet, Vander Linden 1998; prehistoric flint mining in Europe
Collet, Van Neer 2002; Collet, Woodbury 2000; 2001; 2002;
Collet, Vankerkhoven 2004), but both features and material The Spiennes mine offers one of the longest histories of
are being studied. At the same time the area investigated archaeological excavations in Europe, indeed the world,
48
and has played a major role in research into prehistoric velopment of hunting expeditions and transhumance into
flint mining. It was presented at International Congresses something like regular commerce.
of Anthropology and Prehistoric Archaeology in the 19th
and early 20th century, and has always been discussed in In History of Mankind: cultural and scientific develop-
archaeological textbooks (e.g. Mortillet 1885; Cartailhac ment, published under the auspices of UNESCO, in the
1896; Dchelette 1908; Childe 1925). Especially impor- very brief description of mining and trade during the Neo-
tant was the famous Spiennes Section of a railway cutting, lithic, only Spiennes, Grimes Graves and Grand-Pressigny
discovered in 1867, still a reference even in the 20th centu- are mentioned (Hawkes 1965: 431-433), while in World
ry, although archaeological methods and our knowledge of prehistory in new perspectives by G. Clark (1977: 134) we
prehistory have progressed enormously since then (Clark, find only one mention of Grimes Graves.
Piggott 1933; Clark 1952: Fig. 99; Piggott 1954: 39).
During the second half of the 20th century, Grimes Graves
The discovery made by workmen engaged in railway con- gradually took over the place of Spiennes in the reflections
struction in Spiennes had a direct influence on the correct on prehistoric flint mining in Europe. One of the reasons for
interpretation of pits noted in the area of the English-hill this change was the publication, in the 1920s and 1930s, of
fort at Cissbury, W. Greenwells excavations at Grimes A.L. Armstrongs new findings concerning Grimes Graves
Graves, and on the slightly later excavations at Rijckholt- and, in the 1970s, the new, extensive and very professional
St. Geertruid (Lech 1991: 557; 1997: 611-613). Also, in excavations of the site (e.g. Armstrong 1923; 1927; 1934;
later years, the Spiennes excavations inspired systematic Mercer 1981; Longworth et al. 1991; Longworth, Varndell
explorations of other mines. Work at Grimes Graves was 1996). The stronger institutional position of British prehis-
undertaken in the spring of 1914, after the spectacular re- tory when compared to Belgian and its dominance over the
sults of Baron A. de Los excavations were made public field of archaeological syntheses also played a role. At the
(Lo 1913; 1914; Clarke 1915). In 1933 the Spiennes ex- same time, progress was being made with excavations of
cavations were an important factor in dating the British other important sites of prehistoric mining, such as the ones
flint mines (Clark, Piggott 1933). at Krzemionki Opatowskie, Rijckholt-St. Geertruid, Defen-
sola A and Jablines (Weisgerber et al. 1980; Borkowski et
With time, the systematic excavations at Grimes Graves al. 1991; Grooth 1991; Bostyn, Lanchon 1992; Di Lernia et
came to play an increasingly important role in the charac- al. 1995; Felder et al. 1998).
terizations of prehistoric flint mining in Western Europe.
In Prehistoric Europe. The Economic Basis by Grahame In Poland, for example, the first information about the
Clark (1952: 174-178) Grimes Graves served as the ba- discoveries at Spiennes appeared in reports from the In-
sis for a description of prehistoric flint mining in Europe, ternational Congresses of Anthropology and Prehistoric
though Spiennes continued to be much in evidence. For Archaeology in Brussels, written in 1873 by Count J. Za-
Gordon Childe (1957: 293), in his last edition of The wisza, the father of Polish prehistoric research. After the
Dawn of European Civilization, the Spiennes mine was discovery of the mine at Krzemionki Opatowskie, Spiennes
an example of specialization of labour among the clans of became a reference point for the Polish site (Krukowski
Michelsberg villagers: 1939: 49 and 122) and for various descriptions of European
Secondary industry and trade played a recognizable mining (Tabaczyski 1970: 275-276; 1972: 68; Hensel,
part in the Michelsberg economy. Thus at Spiennes in Tabaczyski 1978: 140; Jadewski 1984: 165; Lech 1981:
Belgium lived a community of specialized flint-miners Fig. 2 and Table 2; 1991: 557-561; 1997: 611-623).
skilled at sinking shafts and digging out subterranean
galleries. Indeed, the Michelsberg settlers there consti- Consequently, Spiennes is now considered to be among
tuted a specialized industrial community, supplementing the four or five most prominent sites of prehistoric mining
their livelihood by exporting the products of their mines in Europe, next to the mines at Grimes Graves, Cissbury,
and workshops - and Spiennes was no isolated phenom- Rijckholt-St. Geertruid and Krzemionki Opatowskie (Li-
enon within the Western complex. It implies also the de- chardus et al. 1985: 37 and 441; Grooth 1991: 153; 1997;
49
Whittle 1996: 280; Lech, Leligdowicz 1998: 121 and 125; mining field, on the basis of surface flint distribution. Shafts
Lech, Longworth 2000: 19-20; Milisauskas, Kruk 2002: were noticed a long time ago in the area between the west-
214-215). As a result of many years of farming, as well as ern and southern border of Camp--Cayaux and the Centre
amateur digging and looting in the second half of the 19th de recherches archologiques, previously called the Muse-
and first decades of the 20th century, considerable dam- um (Fig.3.1 and 3; Lo, Munck 1891; Lo, Rahir 1929). In-
age has been done to the parts of the mine lying nearer the dications are less numerous beyond Harmignies Street but
surface, so the Spiennes site is not as well preserved as the several exist (Fig. 3. 16-17 and Lefort, Cody 1953; Lefort
others just mentioned. However, shafts 1 and 2, excavated 1954; Verheyleweghen 1962; Collet, Woodbury in press b).
at the beginning of the 20th century by Baron A. de Lo, Moreover, in the northern part of Camp--Cayaux where
remain the deepest of all known European flint mine shafts meadows hide evidence of flint waste, investigations show
and new excavations have shown the continued research a wider extension of flint extraction than was expected. For
potential of this complex site (e.g. Collet et al. 1997; Tous- instance, about 20 shafts were discovered during the laying
saint et al. 1997; Collet, Van Neer 2002; Defgne, Collet of a gas pipeline in 1992 (Fig. 3.15; Fechner et al. 1993). In
2003; Collet 2004). the northwest sector 17 new shafts were identified during
a rescue excavation in 2005 (Fig. 3.13). Flint exploitation
5. The topography of the mining site is now confirmed as far as the first houses of the village
along Harmignies Street, around the 57m contour line. The
The extent of the exploited areas at Spiennes can be large- presence of mines can be expected along the west border
ly deduced from the mining waste present on the surface. of Camp--Cayaux up to the 52 m contour line, where flint
Typical artefacts are flint nodules and their fragments, exploitation was noticed (Fig. 3.14). On the other hand, in
flakes and other flint waste, along with carbonated de- other parts of the site, there is no reliable indication of flint
posit, and chalk fragments. The latter material can only extraction. Such is the case in the east and northeast part of
be taken into account in those parts of the site where the the mining area of Camp--Cayaux, where only workshops
chalk is covered with a thick layer of silt. In some other were noticed by F. Hubert. In the same way, only the pres-
parts of the site, this kind of evidence cannot be used, as ence of workshops has been determined in the southeast
the chalk naturally appears near the surface. In the north- part of Camp--Cayaux. In the same area however, some
ern part of Camp--Cayaux, prospecting for flint material anomalies were reported during the digging of geological
on the surface was limited by the presence of meadows. trial borings (Fig. 3.d). The geologists thought that they
The extent of flint extraction was also evidenced thanks might have dug through flint shafts (Centre archologique
to the numerous excavations and field observations made de recherche minire 1990: 4.17-4.19).
since the discovery of the site in the nineteenth century. It
is worth noting that the investigations of the last decades In Petit-Spiennes, north of the railway, flint extraction is
have greatly increased our knowledge of flint extraction confirmed by several excavations and some field observa-
in Spiennes (Heim 2003; Fechner et al. 1993; Collet et al. tions (Fig. 3.6 and 3.12; Gosselin 1986; Heim 2003; Col-
1997; Collet 2003a; Collet, Woodbury in press a, b and let et al. 1997; Vandevelde, Hubert 1987; Collet 2003a).
c). A systematic mapping of the scattered information as South of the railway, only some workshop areas have been
yet unpublished was recently undertaken and is still in found. In this case it seems highly probable that the ab-
progress. The documentation collected by F. Hubert and F. sence of evidence is due only to the lack of investigation.
Gosselin and conserved at the Archaeological Service of
the Walloon Region is helpful in this enterprise. In the mining area called Versant de la Wampe, several
mining features are known in the south western part of the
The present state of knowledge concerning the mining Petit-Spiennes plateau, on the upper part of the slope sur-
complex at Spiennes is shown on a map in figure 3. It must rounding the river La Wampe near the 70 m contour line
be clear that this is but a fragmentary view, merely reflect- (Fig. 3.9; Hubert 1976b). One isolated shaft was also iden-
ing the state of research in the different areas. However, tified near the top of the plateau on the same slope between
this mapping generally confirms the physical extent of the the 72.5 and 75 m contour lines (Hubert 1971: 50-51).
50
6. Mining methods
Different mining methods were used at Spiennes, depending

on the geological conditions of the deposits, as noted at oth-
er similar sites, such as Grimes Graves, Rijckholt-St.Geer-
truid and Krzemionki Opatowskie. The simpler exploitation
units are poorly documented. Few are known and only one
was the object of systematic excavations (Verheyleweghen
1953). Others were just explored by chance during the vari-
ous campaigns (De Pauw, Van Overloop 1889-1890; Hubert
1976a; Collet, Woodbury in press c) or only partially exca-
vated (Hubert 1969). Moreover the published data about the
methods are only fragmentary (Verheyleweghen 1953).
On the western border of Camp--Cayaux, remains of ex-

traction and production were excavated on the escarpment
surmounting the river La Trouille. The remains showed
horizontal niches dug straight into the slope. Each one
was filled with a first layer composed of flint waste de-
scribed as workshop, topped by a layer of yellow sedi-
ment mixed with blocks and chalk powder without any
flint remains, either natural or knapped (Fig. 3.5 and 7).
J. Verheyleweghen (1953) interpreted the system of ex-
ploitation as follows. The miners first dug short galleries Fig. 8: Spiennes, Hainaut province. Camp--Cayaux. The slope of the La
about 2.5 m long and 0.7 m high into the slope. When this Trouille river valley. Flint exploitation remains found in 2005: a - modern
embankment; b - natural chalk rock and flint seams; c - filling up deposit,
first extraction and production was exhausted, the roof chalk in powder; d - working floor of hardened chalk; e - filling up de-
posit, chalk in blocks; f - crumbled deposit of natural chalk; g - modern
of the first exploitation unit was pulled down to start a
digging filled by humic sediment; h - slope; i - Camp--Cayaux. Photo-
new one. A gallery dug directly into the slope was discov- graph by M. Woodbury.
ered in the vicinity during building work at the end of the
19th century (Fig. 3.2). Unfortunately L. De Pauw and E.
Van Overloop (1889-1890) focused more on the chrono- Open shafts mining pits are known near the rims of the
logical succession of the workshops connected with this two plateaus, where the silt layer is thin (Briart et al. 1868;
gallery than on the exploitation method. This discovery Breuer 1930; Hubert 1969; 1976a). Some have been ob-
may have suggested his model to J. Verheyleweghen. In served in the cross section of the railway trench, where
2005, a structure with a similar feature was observed in the sediment covering the chalk is only 1 m thick (Fig.
the northern part of the same escarpment (Fig. 3.14; Col- 3.a; Briart et al. 1868). Exploitation by trenches or large
let, Woodbury in press c). It is certain that this feature was pits up to 3 m deep was also described at Camp--Cayaux
dug to extract flint, as the chalk was left behind. The suc- (Fig. 3.4; Breuer 1930). In the 1965 excavation, where the
cessive stages of the exploitation units are covered with chalk appears 1.2 m below the surface, two features were
layers of chalk in blocks separated by several layers of also considered as simple pits (Fig. 3.7). One was exca-
powdery chalk hardened by the circulation of the miners. vated down to 2.5 m, apparently without reaching the bot-
The summit shows a collapse of natural chalk (Fig. 8). tom of the shaft. In the 1975 excavation in Petit-Spiennes,
Unfortunately, dating was impossible, because no frag- three simple pits of 3.5 to 3.9 m depth were discovered
ment of worked flint or other archaeological artefact ap- (Fig. 3.10 and 9). They were dug in a place where the
peared. Therefore the dating of these flint mines remains chalk is about 2.5 m deep. One of them has a preserved
uncertain. side exploitation (niche) 1.3 m long and high. The other
51
Fig. 9: Spiennes, Hainaut province. Petit-Spiennes. 1975 cross section near the railway trench. After F. Hubert (1976a): a - mining pits; b - shafts; c -
galleries.
also shows such a lateral niche but the roof had collapsed In shafts 53.2, 79.3 and 86.8.2 (area 1953), the first seam
and the original state could not be described. was occasionally extracted, while the second and the third
were worked systematically. In shaft 11, explored between
Underground galleries are more fully documented. At Spi- 1997 and 1999, only the first was worked. The ongoing in-
ennes the shafts with underground galleries are between vestigation of shaft 20 also shows a shaft 10 m deep, with
3 to 16m deep. Twelve mines were fully excavated and an exploitation of the first flint seams.
some more are currently being explored. For the rest, only
the mouths of numerous access shafts are known. Under- On the surface, shafts 80.4 and 53.2 are 1.7 m and 2.4 m
ground mine galleries are located mainly in the centre of wide in diameter. Deeper, the same shafts are slightly oval,
the plateau, where this is the only technique appropriate measuring 0.7m x 0.8 m in the case of shaft 80.4, and 1.1
to exploit the flint seams. However, they also occur in ar- m x 1.3 m across in the case of shaft 53.2. Shaft 11 is dif-
eas where flint was exploited by other methods, like the ferent, remaining wide right down to the exploitation level
digging of simple pits, as was recorded by some previous (Fig. 10). Its cross section is irregular with breadth varying
investigations (Fig. 9; Hubert 1969; 1976a), and as can be from 3 m in surface to 2 m at 2 m below and then again 3 m
seen in the cross section of the railway trench (Fig. 5). wide at 3.8 m below and finally 2 m at 5.6 m below. Shaft
20 is even wider. At the surface it is oval shaped, with a di-
At Petit-Spiennes, two zones with underground galleries ameter of 6.2 m x 5.2 m. At a depth of 3.4 m, it is 2.6 x 2.2
have been excavated: the first, since 1953 by the SRPH m wide then at 5.6 m depth it enlarges again to 3.7 x 3.1 m
(Fig. 3. 6; Gosselin 1986), and the second in 1997-2004 wide. In these two cases, the widening results from a dete-
(Fig. 3. 12; Collet, Van Neer 2002). Both areas reveal rioration of the shaft walls after the exploitation ended. In
shafts from 8m to 10 m deep depending on the thickness the case of shaft 20, it seems that the original diameter of
of the sediment. The miners had to dig down to 4.3 m, and the shaft was only a little more than 1 m across.
5.6 m respectively, before they could reach the geological
layers of chalk, due to the natural sloping of the seams. A In the 1953 area, the shaft usually widens one meter below
first flint seam usually appears around one meter below the top of the chalk level and forms a bell shaped room
the chalk. which can reach 3 m high (Fig. 11). From there, the gal-
leries radiate outwards in several directions up to 4 m, or
In these two areas, mines are known in which only the first even to 5.5 m from the centre of the shaft. The height of
flint seams struck upon during the digging were exploited. the galleries varies from 1 m to 2 m. This is due to the fact
52
Fig. 10: Spiennes, Hainaut province. Petit-Spiennes. Shaft 11, excavated in 1997-1999. SW-NE cross section. a: clayey silt; b: loess; c: grey-blue sand;
d: flint gravel, chalk, and sand; e: glauconiferous sand; f: chalk, and glauconiferous sand; g: chalk. Drawing by M. Woodbury.
53
Fig. 11: Spiennes, Hainaut province. Petit-Spiennes. Shaft 53.2. Bell- Fig. 12: Spiennes, Hainaut province. Petit-Spiennes. Shaft 20. Flint picks
shaped room at the base of the shaft. Photograph by J.-L. Dubois. cluster found in the exploitation level of the shaft. Photograph by Y. Gau-
vry.
that in these mines two different flint seams, separated by 1 period of exploitation. In shaft 53.2, dating from the first
m of chalk, were extracted from the walls at the same time. period, Michelsberg sherds from a bottle-shaped vessel
Excavations of shaft 11 in 1997-1999 (Fig. 3.12) showed were discovered in the shaft filling (Gosselin 1986).
short galleries (niches), at times only 0.7 m high (Fig. 10).
Here only one flint seam was exploited. In the different In the centre of Camp--Cayaux, beneath the Centre de
units investigated, the surface worked covers 21 to 25 m. recherches archologiques, there are shafts 15 m to 16 m
In both areas, the extracted flint consists of irregular nod- deep (Fig. 3.3, 14 and 15). The explored underground area
ules from 2.5 cm up to 34 cm long. They were mainly taken covers about 250 m. Like elsewhere at Spiennes, the two
from the walls, and much more rarely from the bottom. The access mouths of the excavated shafts 1 and 2 are about 1
only mining tools found so far in the 1953 SRPH area and m wide. The miners had to dig 5 m to reach the chalk but
shaft 20 (area 1997, Fig. 12) are flint picks. The most com- here they did not exploit the first flint seams and passed
mon are bifacial, thin, with an almond shape. Much less through 15 flint layers before reaching the flint they were
numerous are the irregular and massive multifacial picks seeking (Fig. 15; Hubert 1978: 8, 12). At that depth, a twin
(see below). In the 1997 area, shaft 11 has yielded a mixed flint seam occurs, consisting of large slabs 1 m wide and up
assortment of implements comprising a rich set of red deer to 1-2 m long and about 20 cm thick. Only the lower level
antler tools and massive flint mining tools (Fig. 13). was extracted, while the other was left in situ, apparently in
order to protect the galleries. As at Petit-Spiennes, the bot-
In the area excavated by the SRPH since 1953, at least two tom of the shaft forms a room surrounded by chalk pillars.
different periods of mining were identified. The first dates The mine workings radiate outwards in several directions
from between 4400 and 3700 cal BC and the second from for a distance of up to 4 m. These measurements are esti-
a period between 3500 to 2900 cal BC. Shaft 11, in the mations, as it is sometimes difficult to know which shaft
area excavated since 1997, is connected with the second is connected with the galleries. Some may have even been
54
Fig. 13: Spiennes, Hainaut province. Petit-Spiennes. Shaft 11. Mining tools: 309 - a-b: wear marks, c - undefined parallel notches; 318 a, b, c - wear
marks; 320 and 342: drawings by E. Gumiska; 309 and 318: drawings by P.-Ph. Sartieaux.
55
Fig. 14: Spiennes, Hai-

naut province. Camp--
Cayaux. Shafts 1 and 2.
Plan of the shafts base
with galleries. After F.
Hubert (1978).
Fig. 15: Spiennes, Hainaut province. Camp--

Cayaux. Shaft 2. Cross section. After F. Hubert
(1978).
56
Fig. 16: Spiennes, Hainaut province. The prehistoric mining complex. Radiocarbon dates.
8-9 m long. The height of the levels exploited varies from antler were also found (Bonenfant, Auzou 1993). No ra-
0.7 m to 1 m. Flint was extracted from the roof by digging diocarbon date is available but from the discovery in shaft
the chalk under the twin slabs. The area exploited in each 7, at a depth of 1 to 1.3 m, of a Michelsberg bottle-shaped
of the two, probably fully investigated, mines can be esti- vessel broken in situ, we may conclude that these mines
mated at around 45 m. The mining tools used to dig these also date from the Michelsberg period.
galleries comprise only flint picks of various shapes. No
radiocarbon date is available for these mines. The only ele- So far, the underground exploitation of the mining area
ment of dating is given by fragments of a Michelsberg ce- of the Versant de la Wampe has only been partly exca-
ramic vessel found near the bottom of an unexplored shaft vated (Fig. 3.9). Only some of the mine workings were
located near shaft 1 (Hubert 1978: 38). It is generally ac- explored during the excavation of three trenches which
cepted that these mines were exploited by the Michelsberg had been dug to establish the layout of the ditches of the
communities between 4300 and 3700 cal BC. Michelsberg enclosure. The excavation of the ditches went
In the same area, only 20 m from the limits of the under- down to 2 m. This was enough to identify the remains of
ground galleries belonging to shafts 1 and 2 described mining features: two shafts little more than 1 m wide and
above, five shallow shafts were explored (Fig. 3.11; exca- two galleries with roofs located only 1.3 to 1.7 m under
vation 1990: shafts 2, 4, 5, 6 and 7) over an area of 100 m. the surface.
From centre to centre, the distance between these shafts
was 3 to 5 m. Their depth was only 3 to 3.5 m (Bonenfant, 7. Dating
Auzou 1993; Bonenfant 1993). They were 1 to 2 m wide at
the mouth and 1 m or less at a depth of 1 m. The flint was The few available radiocarbon dates indicate that the min-
exploited from niches 1 m long and 1.2 m high. Flint picks ing field was exploited between 5510 to 4230 BP, then from
were the basic mining tool, though fragments of red deer 4400/4200 to 3350/2900 cal BC (Fig. 16). We dont know
57
if exploitation was continuous or not. The mining activi- of a Michelsberg enclosure on the Petit-Spiennes upland
ties were thus contemporaneous with the Middle Neolithic devoted considerable attention to flint material. But it was
period in Belgium but continued long after. The pottery only from 1981 on, with the excavations at Petit-Spiennes
discovered in the mining areas shows that the site was carried out by SRPH, that all the waste flint material ob-
mainly in use during the Michelsberg period (4300-3700 tained during the excavations was collected. A significant
cal BC). The context of the find shows that this pottery part of the material from Petit-Spiennes was then analysed
was, at times, strictly connected with mining exploitation by F. Gosselin (1986). The published results of her stud-
or at least contemporaneous with it (Colman 1957; Hu- ies, constitute, so far, the best presentation of flint material
bert 1978; Gosselin 1986; Hubert, Soumoy 1993). Some from the mining complex at Spiennes, though they include
Michelsberg potsherds were recently discovered in the only material obtained during excavations of shaft fillings
close context of the chipping floor excavated at Spiennes 53.2 and 80.4 and of underground extraction areas of these
- Lot 51c. The remains of earlier periods are very scarce and several neighbouring shafts at Petit-Spiennes.
and not connected with mines. Those of later periods are
also poorly documented. Most potsherds from these later In 1997, H. Collet began excavations at Spiennes, pay-
occupations are not assigned to specific culture and can ing particular attention to the flint material. From 1997 to
date from the Late Neolithic to the Bronze Age. Some Late 2004 she concentrated on the Petit-Spiennes shafts, but in
Bronze Age and Iron Age finds were clearly identified at 2005, while doing some short-term rescue work, she ex-
the mining sites at Camp--Cayaux and Petit-Spiennes cavated large areas of flint workshops at the north edge
(Marin 1961). It remains unclear whether a direct or in- of the Camp--Cayaux mining field (Collet, Woodbury in
direct relationship with mining exploitation can be estab- press a).
lished. Nevertheless, according to F. Hubert (1976c), one
of the Late Bronze Age features is connected with flint 8.1. Flint finds from Spiennes
processing. So far however, incontrovertible evidence is
lacking for mining later than the end of the 4th millennium The results of the Spiennes mining complex excavations
cal BC. completed so far make certain generalisations possible
regarding the basic trends of flint production and the cat-
8. Study of flint working at the Spiennes egories of artefacts. However, little can be said about the
mining field flint workshops and the techniques or technology of flint
working, which left behind a great mass of flakes and flint
Not much is known about flint working in the Neolithic waste. We have at our disposal only some very modest data
Spiennes, as compared to such prehistoric flint mining on the flint inventories, obtained from F. Huberts (1969;
sites as Grimes Graves in England, several of the Polish 1971) excavations, and the recent SRPH research at Petit-
mines or Jablines in France. Although the first articles de- Spiennes.
voted, to some extent, to flint working at Spiennes were
written as early as the 19th century, in connection with the The flint mined at Spiennes served to make blade blanks
research done by L. De Pauw and E. Van Overloop (1889- from cores, and a small number of flake blanks. The pro-
1890), archaeologists of later periods showed little interest duction of axe blades is especially well substantiated and,
in the flint workshops before the middle of the 1960s. to a much lesser degree, that of other core tools, of which
chisels ciseau though very rare, are a typologically dis-
F. Hubert was the first archaeologist to display more inter- tinct category (see Brzillon 1971: 195, 196). Mining tools
est in the flint finds at Spiennes. In 1965, he excavated form a separate group. Picks predominate, while other min-
four workshops, but published only their general charac- ing tools consist of various failed specimens obtained in
teristics, paying greater attention only to the typology of the process of producing various implements, such as axe
the basic categories of artefacts: cores, blades, axes, chis- blades, and then adapted for the job of flint extraction as
els, picks and tranchets (Hubert 1969). Compared to ear- the need arose (Lech 1982-1983: 24-26; 1991: 560-569).
lier works, F. Huberts article (1971) on the excavations
58
Flake cores seem to be a rarer category at Spiennes than

blade cores. The specimens are varied and difficult to at-
tribute to standardised categories. Excavations so far have
turned up flake cores more often at Petit-Spiennes than at
Fig. 17: Spiennes, Hainaut
province. Camp--Cayaux. Camp--Cayaux (Gosselin 1986: 90-91, 114-115).
Flint core for blades from
workshop 7. RBINS: Ste-
vens collection. Drawing by Blades and blade fragments occurred frequently at Camp-
A. Hauzeur. -Cayaux and much more rarely at Petit-Spiennes. The
longest regularly shaped blades were obtained from cores
prepared from flint extracted from the deep shafts of
Camp--Cayaux. Both regular and failed specimens occur.
Cores and blades

Single platform blade cores are among the most important
categories of finds from the mining complex at Spiennes
(Fig. 17). Many of them have standardised features and
could be associated with those workshops which exploited
flint from the deepest shafts. The largest regular stand-
ardised cores would have been prepared from fragments
of flint blocks extracted from the deep shafts at Camp--
Cayaux. They are similar to the famous livres de beurre
from Grand-Pressigny. Compared to the classic livres
de beurre, the regular cores from Spiennes are smaller
and less carefully prepared; also the variety and range of
Fig. 18: Spiennes, Hainaut province. Camp--Cayaux. Blades. RBINS:
flint knapping techniques was more limited. It seems that Stevens collection. Fig 18 a-b: drawings by A. Hauzeur; c-d: drawings
more blades were obtained in the process. After extraction by A.-M. Wittek.
some of the cores were immediately converted into axes.

Core knapping was preceded by preparation of the pre- Measuring a sample of blade scars on such cores from Ch.
core. During this process, much attention was paid to the Stevens collection at the RBINS (Fig. 18), it was deter-
preparation of the blades striking surface. The trimming mined that most of the blades were between 15-21 cm in
technique was generally used. The core exploitation meth- length, 25-35 mm wide and 7-10 mm thick. The flint ma-
ods used at Spiennes were the same as at Grand-Pressigny, terial at Camp--Cayaux was also characterised by large
and at workshops 1 and 3/1960 in Sspw (Dzieduszycka- core trimming blades and further blades with traces of core
Machnikowa, Lech 1976). Blades were struck off from preparation (see also Vanmontfort et al in press).
the partly rounded striking surface until it became flat, at
which point it had to be partly rounded again. In the Petit-Spiennes area, regular blades like the ones
from Camp--Cayaux are rare. Drawings of cores from F.
At the Spiennes mine complex there were areas where Gosselins publication (1986: 92, Fig. 31) confirm the pro-
blade cores were prepared and worked, for instance near duction of blade blanks of up to 20 cm in length.
shafts 1 and 2; and also in the southwest part of Camp--
Cayaux, and at Petit-Spiennes, near shafts 20, 28 and 53.2. Axe blades and chisels
We also know of areas where no cores were found, such as Among specimens that underwent preliminary working,
the north-northwest part of Camp--Cayaux, where rescue early roughouts, and roughouts axe blades were the most
excavations were carried out in 2005. common. All axe blades were bifacial. An analysis of the
59
prepared using the same technique as with axes but slim-

mer in shape (Brzillon 1971: 195 and 196). Chisels were
probably among the tools used during activities carried out
within the mining field area, and altogether they comprise
only a small part of the artefacts produced at the Spiennes
workshops. Another very rare tool at Spiennes is the tran-
chet, a more typical tool in Middle Neolithic domestic
context.
Mining tools: picks and others

The most important mining tool at Spiennes was the pick,
used primarily for digging chalk. Excavations have yield-
Fig. 19: Spiennes, Hainaut province. Camp--Cayaux. Two axe blades

from the area of the Centre de recherches archologiques. Photograph
by M. Woodbury.
roughouts earliest steps and of the roughouts themselves

indicates large variations in the group, both in morphology
and size. The largest known roughout of an axe blade from
Spiennes comes from Camp--Cayaux; it was published by
A. de Lo (1914). This unusual specimen is 28 cm long and
was carefully prepared to the stage which precedes polish-
ing. It was found in or just next to a flint knapping work-
shop excavated in 1913, located near shafts 1 and 2. Two
finished roughouts of axe blades of about 20 cm (Fig. 19)
were also recovered in the same area in 2006. Both speci-
mens were made of massive flakes struck off from large
flint nodules which had been extracted from deep shafts in
this area of the mining field.
During excavations of the southwest part of Camp--

Cayaux carried out in 1965, F. Hubert (1969) found 62 ear-
ly roughouts and roughouts of axe blades and fragments
(he also found 118 blade cores and 17 flake cores). The
biggest is nearly 22 cm long. Large axe blades were also
produced at Petit-Spiennes (Fig. 20). Because of the dif-
ference in the character of the raw material exploited here,
they were smaller than the largest specimens at Camp--
Cayaux.
Among the flint tools from Spiennes we also distinguish

Fig. 20: Spiennes, Hainaut province. Petit-Spiennes. Shaft 11.
chisels or ciseaux. Generally speaking, these are tools Axe blade roughouts. Drawings by E. Gumiska.
60
ed many tools belonging to this category; most of them

found in shafts 1 and 2 at Camp--Cayaux during excava-
tions carried out by Baron A. de Lo. In his first campaign,
about one thousand picks were found (Fig. 21) and in the
second another 500 (Lo 1913: 45; 1914: 35).
F. Hubert (1978: 24-31) analysed 117 picks found during

the excavation of underground galleries between shafts 1
and 2 conducted by M. Lefort in 1948-1954. He divided
them into 7 different types. The most common were fusi-
form picks (Pics fusels typiques) 45 specimens, that is
48.9%. F. Gosselin (1986: 54-87) examined 289 picks from
three shafts in the Petit-Spiennes area: 53.2, 79.3 and 80.4.
She divided them into two main categories: the multifa-
cial (multifacial) including 48 specimens, and the bifacial
(bifacial) comprising 241 artefacts. The latter is the most
common in this area. In shaft 20, whose exploitation level
was partially excavated, about one hundred almost-whole
picks, 35 large fragments, and several hundreds small
chips were distinguished. Almost all of them, following F.
Gosselins typology, were classified as bifacial. Some of
them were only functional picks, i.e. failed early rough-
outs and roughouts of axe blades used in mining work.
The large numbers of picks recovered from the mine pits

and shaft fillings show how important this tool was for
Fig. 21: Spiennes, Hainaut province. Camp--Cayaux. A: Flint pick from the flint miners at Spiennes. These numbers also suggest
workshop 7. RBINS: Stevens collection. B: Flint picks from 1912-1914
excavation of Shafts 1 and 2. A: Photo E. Dewamme, RBINS; B: After that picks wore off relatively quickly and had to be fre-
A. de Lo (1925).
quently replaced by new ones. For example, a broken pick
and fragments were found at the bottom of the backfill in
shaft 20. The refitting of these artefacts gives a good idea
of how tools were progressively worn away during work.
It shows that the fragments came off bit by bit, and not all
at once in one strike (Fig. 22).
Fig. 22: Spiennes, Hainaut province. Petit-Spiennes. Shaft 20. Refitted
bifacial flint pick. Photograph by M. Woodbury.
61
Other flint tools were also used in the process of mining. Nevertheless, their occurrence at Spiennes is certain.
Usually these were previously-discarded roughouts of axe Another unusual category of finds connected with flint
blades. They were adapted ad hoc for the job of mining knapping are bone punchers, used for indirect percussion
in chalk. Evidence of transforming exhausted blade cores to remove blades from cores. They are known from older
into ad hoc functional picks must also be noticed. excavations, and F. Hubert (1969: 28) also found a deer
antler puncher when excavating in 1965 workshop II. F.
8.2. Flint knapping workshops Huberts excavations in 1965 represent the only extended
studies of flint workshops in the history of the Spiennes
Among the tools used for flint knapping at Spiennes were excavations. He completely excavated workshops I and III,
sandstone and quartzite hammers, though flint hammers as well as parts of workshops II and IV, and published the
are equally numerous. This group of tools is extremely general results of his work (Hubert 1969). In workshops I,
variable in size and weight, with two specimens exceeding II, and III he found sherds of Michelsberg culture pottery.
0.5 kg and many considerably lighter ones. Only flint tools and objects made from other raw materi-
als were published from these excavations. Large samples
Another characteristic category of finds connected with the of archaeological material from several other workshops
final phase of making axe blades and chisels are polishing were obtained recently by H. Collet.
stones or polissoirs. 49 such objects have been reported,
not counting the specimens noted in the Michelsberg en- 8.3. Initial analysis of flake samples from Lot 51C, fea-
closure excavations and in the recent excavations (Collet ture 5
2000). Most come from the Camp--Cayaux mining field.
Polissoirs are present - mostly in fragments - but we dont Towards the end of January and during February, 2005,
know about their proportional representation. The largest rescue excavations were carried out on a building lot ly-
comes from Camp--Cayaux and was found in the 19th ing in the north-northwest part of Camp--Cayaux. The
century during some unidentified digging. It measures 41 digging of the 150 m2 trench lasted three weeks, during
cm long, 24.3 cm wide, 29.9 cm thick and offers four sur- which time nearly 50 percent only of the features and ma-
faces for axe polishing (Delvaux 1885-1886: 197). terials were explored (Collet, Woodbury in press a). This
revealed the presence of in situ workshop places. Under
Another tool connected with flint workshops are retouch- emergency conditions, all the flint from the excavated ar-
ers. They are a rare find and it is not always possible to eas was removed en bloc, packed with the sediment to be
determine whether specific traces of utilisation may not sieved later. After it had been collected, 21 features were
be the result of the incidental use, as a makeshift hammer. located, of which 17 were definitely shafts (Fig. 23).
Fig. 23: Spiennes, Hainaut province. Camp--Cayaux. Plot 51c. Plan of the 2005 excavation: a - excavated area; b - distribution of flint material after
removal of topsoil and cleaning; c - house area and parking area in front of house; d - shafts. Drawing by M. Woodbury and A. Hauzeur.
62
An initial study of 1233 flakes and flake fragments was

carried out with preserved butts > 20 mm, together weigh-
ing 4985 g (Fig. 24). The sample was organized into di-
mensional classes, taking into consideration the maximum
size of flakes (Fig. 25), following the divisions adopted
in the classification of workshop material from Sspw
and Grimes Graves (Lech, Longworth 2000; 2006). Inde-
pendently, the specimens were also divided according to
thickness (Fig. 26), as proposed by A. Augereau (1995).
This is the first, preliminary comparison of the results of
two classifications carried out on the same, small sample. A
comparison of diagrams from Fig. 25 and 26 shows much
greater variations in the frequency of specimens in the three
size categories of large, 80 mm, medium, 50-80 mm, and
small 20-50 mm flakes than in the three thickness catego-
ries, i.e. very thick, > 15 mm, thick, 5-15 mm and thin, <
5 mm flakes. As mentioned earlier, from Lot 51c we only
know of workshops preparing flint axe blades and there are
Fig. 24: Spiennes, Hainaut province. Camp--Cayaux. Plot 51c. Sample
of the material explored during the 2005 excavation. Differentiation of no indications that blade cores were prepared there. Flake
size. A - flakes; B - sieved chips. Photograph by M. Woodbury. cores, if they occurred, seem rare.

province. Camp--Cayaux.
Plot 51c. A sample of flint
material from workshop
5/2005. Differentiation of
basic attributes of flakes >
20 mm on basis of the size
(large > 8 cm, middle 5-8
cm, small 2-5 cm) and up-
per surface of specimens:
A total cortical flakes;
B secondary flakes (par-
tially cortical); C non-
cortical flakes.
63

province. Camp--Cayaux.
Plot 51c. A sample of flint ma-
terial from workshop 5/2005.
Differentiation of basic at-
tributes of flakes > 20 mm on
the basis of the thickness and
upper surface of specimens.
(very thick > 15mm, thick
5-15mm, thin < 5mm).
Large Middle Sm all The sample is notable for the predominance of small flakes
flakes flakes flakes
80 mm 50-80 mm 20-50 mm
of 2 to 5 cm from each category (N=1138), totalling 92.3%
of all flakes (Fig. 25). However, with regard to weight their
Cortical flakes 55 g ? 20.6 g ? 3.3 g
share is much more modest 52.4%. At the same time,
though not as evidently, there is a predominance of thin
Secondary flakes 113.2 g ? 22.3 g ? 4.1 g
flakes when it comes to numbers - 76.5% of the total -,
though they constitute only 32.6% in regard to weight. It is
Non-cortical 41.7 g ? 17.4 g 2g the opposite case with large flakes, only 13 (1%) in number
flakes
but they constitute 17% in terms of weight, and very thick
flakes, altogether 21 (1.7%) in number but they constitute
Fig. 27: Spiennes, Hainaut province. Camp--Cayaux. Plot 51c. A sam-
ple of flint material from workshop 5/2005. Average weight of basic cat- 18.7% in terms of weight (Fig. 26). The predominance of
egories of flakes > 20 mm on the basis of the size and upper surface of
specimens.
the whole non-cortical flakes (N=1005) when compared
with the total of cortical flakes (N=91) is only slightly
Very t hick Thick Thin
smaller in number - 81.5% for 7.4% - and still clear, though
flakes flakes flakes
> 15 m m 5-15 mm < 5 mm lesser in the case of weight 64.6% for 9.6%. Generally,
Cortical flakes 28.7 g ? 8.8 g ? 1.6 g
the number and weight analyses confirm the relations de-
termined earlier for the Grimes Graves mine. The analysed
sample was too small to warrant an evaluation and compar-
Secondary flakes 13.7 g ? 9.2 g 2.5 g
ison with the other weight averages of large, and medium
sized flakes, and the very thick flakes. Moreover, studies
Non-cortical 40.5 g ? 9g 1.7 g to determine how both these features are correlated (Fig.
flakes
25-28) have still to be conducted. However, it should be
Fig. 28: Spiennes, Hainaut province. Camp--Cayaux. Plot 51c. A sam- noted that the average weight of a specimen from the group
ple of flint material from workshop 5/2005. Average weight of basic cat-
of partially cortical flakes (9.4 g) is nearly twice the aver-
egories of flakes > 20 mm on the basis of the thickness and upper surface
of specimens. age weight of a totally cortical flake (5.3 g) and even larger
64
but seems to support the argument just presented, which

is based mainly on an analysis of material from workshop
complexes at the Grimes Graves mine (Lech, Longworth
2000; 2006).
8.4. Organisation of the flint working process
Studies of the mine complex at Spiennes and of flint work-

ing in communities of the Michelsberg culture in the Spi-
ennes region and neighbouring areas, allow us to draw only
a preliminary sketch and some general conclusions about
the organisation of flint working here, conclusions much
more modest than those deriving from the Grimes Graves
mine. At Spiennes, flint nodules were extracted from vari-
ous levels and differed in size and quality. In the case of
the recently excavated shafts at Petit-Spiennes, which
provide fresh observations, it was determined that the ex-
Fig. 29: Spiennes, Hainaut province. Petit-Spiennes. Shaft 20. A sam-
ple of flint nodules abandoned by miners in the exploitation level. Note tracted nodules were cleared roughly of chalk, checked for
traces of removed flakes on some nodules. Photograph by J. Lech.
quality by removing the largest protuberances and simple
flakes (Fig. 29). The good nodules were transported to the
than the average weight of a non-cortical flake (3.4 g). surface; the rejected ones were left underground. Among
Given these preliminary considerations, it would seem that the latter were all the small nodules (Fig. 30). The ana-
the first flakes detached from the blank (nodules or large lysed sample of flint material comes from the exploration
flakes) were intended to peel off the surface rather than to of a 15m waste dump left at the bottom of shaft 20 at
check its quality, and to prepare it for the basic
procedures of the following shaping process.
Therefore, these flakes were neither the largest
nor the heaviest.
When the nodules, or their large fragments

were being shaped, the flakes removed were
larger than in the first phase. These are sec-
ondary flakes, with partially cortical surfaces.
During the last phase, when the early rough-
outs, and the roughouts of axe blades, were
being formed, many small flakes and micro-
flakes (with a maximum size of 10 to 15 mm)
were produced; the last two categories were
produced in large numbers during the final
shaping of the edges of the axe roughout.
The sample examined here remains very small
Fig. 30: Spiennes, Hainaut province. Petit-Spiennes.

Shaft 20. Size and weight of flint nodules abandoned
by miners in the exploitation level.
65
Fig. 31: Spiennes, Hainaut province. Petit-Spiennes. Shaft 20. Differentiation of basic categories of flint waste aban-
doned in the exploitation level, according to frequency and weight of specimens.
Fig. 32: Spiennes, Hainaut province. Petit-Spiennes. Shaft 20. Differentiation of basic categories of flakes abandoned in the
exploitation level, according to thickness and upper surface of specimens.
66
Petit-Spiennes, constituting about 25 per cent of the total The occurrence of non-cortical flakes resulting from the
mass of dumps. The sample seems accurately to reflect the processing of the extracted nodules is in fact much lower
structure of the whole dump (Fig. 30-31). However, the than suggested by the diagram (Fig. 33: C and F), as part of
analysis disregards microflakes and chips < 15 mm, to- the flakes in this category are the result of damage to and
gether weighing 18.73 kg, and numerous flint picks which repair of picks. In the dump at the bottom of shaft 20, there
will be analysed in the future. is a large number of waste pieces.
Interestingly, the sample did not contain flint hammers and To describe the flakes from the waste dump at the bottom
only two sandstone hammers were found in the whole area of shaft 20, H. Collet used the method proposed by A. Au-
explored in the shaft. Most of the hammers were probably gereau (1995) but supplemented it by counting the classi-
removed from the shaft by the miners. fied flakes (Fig. 33: A-C), though not the specimens in the
It is worth noting that many of the flint nodules extracted

did not meet the miners requirements, because of their
size or shape, and were discarded straight away and left
underground. A large number of such nodules were re-
covered 763 specimens, constituting in weight over two
thirds of all the flint left at the bottom (Fig. 30). The nod-
ules were widespread throughout the dump, indicating that
the nodule was assessed for suitability and, if needed, was
disqualified immediately after extraction from the chalk.
Over half of these discarded nodules weighed less than 1
kg. 38% were 10-15 cm in size and it is obvious that small
specimens did not interest the miners; the same is prob-
ably true for the slightly bigger ones, of 15-20 cm (24%).
We can assume that most of the nodules which were taken
above ground were larger than 20 cm and weighed more
than 2.5 kg (Fig. 30). The relatively high percentage of Fig. 33: Spiennes, Hainaut province. Petit-Spiennes. Shaft 20. A speci-
men from the horn category abandoned in the exploitation level. Photo-
defective nodules means that the quality checks and the graph by M. Woodbury.
selection carried out at the bottom of the shaft were accu-
rate. The relatively large percentage of nodule fragments chip category (<15 mm), which were only weighed. This
- beside natural ones - indicates that damage also occurred allows the interesting possibility of comparing the number
during the extraction and the checking process. of pieces in the different categories of flakes with their
weight and of evaluating the significance of the criterion
At the same time as they were checked for quality, any kind of thickness, its correlation with the features of the dorsal
of irregular protuberance was removed from the nodules. face of the specimens and their weight (Fig. 33).
These horns (418 pieces 3.64%) were probably also
perceived by the miners as diminishing the pieces worth The numbers of totally cortical, partially-cortical and
(Fig. 30 and 32). Both procedures were complementary. non-cortical flakes in the dump at the bottom of shaft 20
Quality was also checked by striking off simple flakes, were similar 31.95%, 34.21% and 33.84% respectively,
mainly cortical and secondary, and these two categories but in terms of weight, they were respectively: 43.35%,
together predominate in number (66% of all flakes) and 45.54%, and 11.11% - so we see a clear difference in the
in weight (88.9%) (Fig. 33). Among the flint workshops mass of non-cortical flakes when compared with the other
from various European mines which have been studied so two groups. The differences in the thickness of flakes are
far, similar proportions have been determined at Sspw, even more interesting. Here the relation between the dis-
in the mine workshops located on the surface (Dzieduszy- crete categories of thickness shows clear differences and,
cka-Machnikowa, Lech 1976: 92-100; Lech 1983: 52-64). moreover, the ratio of number to mass of flakes is inverse-
67
ly proportional. In number, flakes > 15 mm thick made up A. Augereau seem pertinent, but there is no support for the
12.06% of those found at the bottom of the shaft, flakes assumption that in the case of flakes these criteria function
between 5 and 15 mm made up 38.84%, and flakes < 5 more effectively than the criterion of size.
mm, 49.1%. In terms of mass the same categories consti-
tuted: 61.03%, 30.52% and 8.45%, so the infrequent very Concerning flint mining from the deepest shafts, such
thick flakes were heaviest, while the most numerous cat- as shafts 1 and 2 at Camp--Cayaux, the organisation of
egory of thin flakes was the lightest. Thus we can regard work was completely different. The largest nodules were
the thickness of flakes as an important criterion for classi- extracted there, from 1 to 2 m long, about 1 m wide and
fication, in accordance with the postulates of A. Augereau 15 to 30 cm thick. The nodules were split up underground
(1995: 146) and J. Pelegrin (1995: 162), and in accordance and the raw material was brought up in the form of smaller
with their place in the chane opratoire. blocks which were small and light enough to be transport-
ed from a considerable depth to the surface through narrow
If we consider the differences in the flakes from the dump shafts. Probably part of the waste material was removed
at the bottom of shaft 20 according to their connection with natural pieces of the flint nodule, flint breccia, sparse
successive elements of the chane opratoire, 25.83% of detached flakes, etc.
the specimens, in regard to numbers, and 71.06%, in regard
to weight, should be attributed to the stage of preparation Unfortunately, there are no observations which would help
of the nodule when very thick, and thick cortical flakes and describe the behaviour of miners in the direct vicinity of the
very thick partially-cortical flakes, were detached by means shafts. This is due to the fact that, at Spiennes, at no time
of a hard hammer. This is an acceptable conclusion. were the dumps next to the mine shafts, or rather their mod-
est remains preserved in some cases by the shafts, studied.
According to A. Augereau, flakes of medium thickness (5 Nor was there any research directed towards excavating the
to 15 mm), partially-cortical and non-cortical should be surface area surrounding the shafts which, in some cases,
attributed to the stage when the nodule was being shaped may have survived.
using a hard or soft hammer. This would apply to 24.16%
of the specimens with regard to number and 18.3% with The raw materials were transported from the shafts to work-
regard to weight. The problem is that there are no indica- shops located within the mining field. In workshops of the
tions that axes were prepared at the bottom of the shaft. Michelsberg culture, blade cores were prepared and exploit-
It seems more likely that some of the flakes in this group ed, while part of the raw materials were directly transformed
are the result of incidental damage to the nodules when into flint axe blades; in some cases, (if suitable), used blade
they were being extracted from the deposit with the use cores were turned into axes. At times, the numerous Spi-
of picks. They could also have been the result of removal ennes picks were produced at the same time and in the same
of flint protuberances from the nodule, the testing of the workshops as manufactured blade blanks and axe blades.
quality of the raw material and the forming and repairing This is true in the case of workshops II, III and IV, excavated
of picks, or of damage to the picks sustained during work. by F. Hubert in 1965. Picks were prepared there from used
blade cores (Hubert 1969: 27, 28 and 31). We know that
In contrast, most of the 25.14% of thin non-cortical flakes in the Spiennes area there were also workshops where only
of 5 mm (probably small, considering their weight which axes were made, while no blades were produced.
was only 3.40% of all the flakes) and part of the uncounted
small fraction of microflakes and chips of 15 mm and Workshop productivity is difficult to determine. Only
mass of 18.730 kg, were most probably the result of the on the basis of F. Huberts excavation from 1965 can we
final stage of preparing picks and other mining tools, of venture a guess that in the case of workshops I IV the
repairing them, and most certainly also of damage done scale of production was not large. It probably comprised
them during work. blades procured from about 100 cores, a similar number of
axes and a substantially smaller number of picks. There is
All in all, the criteria of thickness and weight proposed by a large margin of error here. We cannot exclude the pos-
68
sibility that the same workshops also produced advanced 9. The Petit-Spiennes enclosure and
pre-cores and prepared cores which, together with selected remains of daily life
blades and flakes, roughouts of axe blades and initially
prepared flint blocks, as raw material, were transported to The Spiennes mining complex also included a settlement
the miners own settlements or became goods for exchange surrounded by a system of ditches and banks, considered
with other communities. This possibility is indicated by the by some authors as to have been a place serving some cer-
results of excavations at the settlement of a Michelsberg emonial function.
culture community in Thieusies (Vermeersch et al. 1990) Discovered in the fifties, the enclosure is located on the
at less than 15 km from the flint mines. Four flint picks Petit-Spiennes plateau south of the railway trench, on a
found at the settlement suggest that the Thieusies commu- site where some mining features were discovered, and in
nity was directly engaged in exploiting the Spiennes flint. the immediate vicinity of the mining area of Petit-Spi-
The large number of cores in Spiennes flint, as well as the ennes. The enclosed area, labelled Camp Michelsberg, is
presence of axe roughouts, also supports this hypothesis. situated near the top of the plateau and looks down upon
Unfortunately, our great ignorance regarding settlements the surrounding countryside. Archaeological excavations
of the Michelsberg culture communities in Hainaut prov- determined the location of ditches and only one related pit
ince, in all of Belgium and the neighbouring regions of was excavated inside. As a result, a lot of flint flakes and
France, makes further surmises as to the connections be- some burnt daub, sandstone fragments, potsherds, burnt
tween the Spiennes mine complex and its settlement quite bones as well as charcoal were found.
pointless.
Fig. 34: Spiennes, Hainaut province. Petit-Spiennes. Michelsberg enclosure on the Petit-Spiennes plateau: a - interruption; b - excavation trenches.
Drawing by M. Woodbury after F. Hubert (1971 and 1976a).
69
The layout of the ditches and the approximate size of the

enclosure were determined only after some thirty trench-
es had been dug, though this was not enough to confirm
whether the ditches were continuous or not. Only one at-
tested interruption was discovered on the eastern part of
the enclosure. For this reason, the drawing set out here
(Fig. 34) gives only a general outline, based on an aerial
view and on the excavation data published up to 1976
(Hubert 1971; 1976a).
The feature is composed of two parallel ditches 2.55 to 8 m Fig. 35: Spiennes, Hainaut province. Camp--Cayaux. Flint implements.
wide and 1.3 to 2.45 m deep beneath the present day sur- RBINS: Stevens collection. Fig. 35 a: drawing by A.-M. Wittek; Fig. 35
b-c: drawings by A. Hauzeur.
face and enclosing an area of about 14 ha. The distance be-
tween the two ditches varies from 5 to 10 m. The presence,
originally, of two banks on the inner side of each ditch was was first studied by F. Hubert (1971), and recently by B.
suggested by F. Hubert (1971), based on the distance be- Vanmontfort (2004). The archaeological material can be
tween the two ditches and an examination of their fills. attributed to the Michelsberg culture (Hubert 1971) and,
specifically, to the Central Scheldt basin group (Van-
Among the flint material from small excavation trenches, montfort 2004: 342). This group is connected with the
F. Hubert (1971: 42) distinguished flake and blade cores Michelsberg culture groups from the Paris basin, the Mid-
the former in greater numbers. The most frequent tools dle Meuse and the Rhine basin. This material shows no
were flake end-scrapers and axe roughouts; the latter were real difference when compared with the pottery found in
three times more numerous than polished axes. There were the mining areas: the same temper, same treatment, same
also chisels, side-scrapers, flint arrowheads, and a single shapes (Vanmontfort 2004: 214) For this reason the set-
burin and borer. The connection of the settlement with the tlement looks contemporaneous with at least a part of the
mining and knapping of flint is supported by the presence mining activity at Spiennes.
of picks, hammers, polishing stones (polissoirs).
The pollen spectrum obtained from three samples taken in Elsewhere, the Spiennes mining fields were dotted with
one profile, located in the deepest level of the inner ditch, structures connected with the exploitation and working
shows that the enclosure was situated in an area which of the flint. Excavations revealed Michelsberg potsherds,
was probably largely cleared of forest. The arboreal pol- bones of consumed animals and domestic flint tools, gen-
len represents only 19 to 25 % of the total number of the erally rare at Spiennes. End-scrapers, burins, blades and
pollen spectrum. It contrasts significantly with the results polished axes were found (Fig. 35). Unfortunately, again
obtained for the mining field itself, which shows a great- no clear interpretation can be given of these discoveries,
er quantity of trees and mainly bushes - from 42 to 95% which are generally old and insufficiently recorded.
(Defgne, Collet 2003; Heim 2003). The human activity is
also attested by the presence of cereals and ruderal plants 10. Human bones remains
which clearly indicate the presence of fields and grazing in
the immediate vicinity (Heim 1971). It seems that most of the human bones discovered at Spi-
ennes and in its vicinity give evidence of funerary prac-
Unfortunately, no radiocarbon dates are, as yet, available. tices from the Middle Neolithic period, other than those
The archaeological material suggesting a date for the en- found during the same period in cave or rock-shelters in
closure comes mainly from the ditches, especially from the the Meuse basin (Cauwe et al. 2001).
entrance and the inner ditch. The pottery from the Middle
Neolithic layers (it should be pointed out that material of Human remains are known from Obourg, Strpy and Spi-
later periods was found in the upper part of the ditches) ennes. Some of these have been attributed to miners acci-
70
dentally killed during their work, like the famous Obourg but this seems an exceedingly modest estimate.
miner. Some 15 years ago, most of these human bones As to the intriguing problem of labour specialisation, it
were studied in detail, also in the historical context of seems much more likely that flint mining was just one of the
their discovery, and partly dated (Heinzelin et al. 1993). In spheres of activity of farming communities who exploited
fact, some of these human remains turned out to be forger- the mines and at the same time worked the land and bred
ies, dating from the turn of the 20th century. The forgers animals (Gautier, Biondi 1993; Collet 2004: 132). These
used human remains from the Merovingian period as well modest conclusions show the crucial problems which ar-
as protohistoric bones, as in the case of the miners of chaeology encounters to interpret such sites as Spiennes
Obourg and Strpy. and other nearby mines. The key to solving them seems to
lie in comparative studies of mining fields, flint workshops
Nevertheless, the results of different analyses provided and flint material from settlements (Dzieduszycka-Mach-
some information about the presence of human remains in nikowa, Lech 1976; Saville 1981; Lech 1983; 1997: 624-
a mining context. 14C dates indicate that a skeleton from 628; 2004: 70-76; Augereau 1995; Cupillard et al. 1995;
Spiennes has been correctly dated to the Middle Neolithic Jeudy et al. 1995; Pelegrin 1995; Lech, Longworth 2006).
(Spiennes C; OxA-3196; Fig. 16). The most complete and
certain find is a female adult with a perinatal child placed Brussels Warsaw Mons Paris, June 2005 September
in the middle part of the filling sequence of shaft 11 at 2007
Petit-Spiennes (Toussaint et al. 1997; Collet, Toussaint
1998; Collet, Van Neer 2002). A 14C date gives 4500 50
BP (Beta-110683), which is within the range of dates for Hlne Collet
the flint mining at Spiennes (Fig. 16). Association sans but lucratif Socit de Recherche
prhistorique en Hainaut
Final remarks Maison Losseau
37, rue de Nimy
On the basis of the excavations carried out so far, it is diffi- B-7000 Mons (Belgium)
cult to determine how much mining and flint working was e-mail: hcollet@tvcablenet.be
done overall in the Spiennes mining fields (Collet 2004:
132). We do not know the exact borders of the mining Dr Anne Hauzeur
fields and therefore cannot calculate the number of min- INRAP Centre - le-de-France, scientific collaborator at
ing features. There is also no precise dating for the vari- RBINS
ous types of mining activities. Moreover, it is not known 148, Avenue Andr Maginot
whether the flint mining at Spiennes was continuous, or F-37000 Tours (France)
whether the different phases of exploitation of the flint de- e-mail: ahauzeur@yahoo.fr
posits were divided by time intervals when mining activity
ceased. Prof. Dr Jacek Lech
Polish Academy of Sciences
The size of the Spiennes site complex and the advanced Institute of Archaeology and Ethnology
technology of the extraction process led some authors to Aleja. Solidarnoci 105
infer that the mining and working of the flint there was PL 00-140 Warszawa (Poland)
highly specialised (Verheyleweghen 1966). Such conclu- e-mail: lech@iaepan.edu.pl
sions need to be verified. For reasons mentioned earlier,
we know little about the scale and extent of distribution
of products from the workshops located next to the mines.
According to available data and literature, there is no doubt
that the blades and axe blades made in the mine workshops
were distributed within a radius of 70 km from the mine,
71
References Breuer, J. 1930. Recherches Spiennes 3-20 mars 1930.

Bulletin des Muses royaux dArt et dHistoire 3, 87-88.
Adam, A. 1959. Sur quelques outils facis dhabitat
trouvs Spiennes (Pa dla liau). Annales du Cercle Brzillon, M. 1971. La dnomination des objets de pierre
archologique de Mons 63, 63-66, 4 pl. taille. Matriaux pour un vocabulaire des prhistoriens de
langue franaise. IVe suppl. to Gallia Prhistoire. 2nd ed.
Armstrong, A.L. 1923. Discovery of a new phase of early Paris: Editions du CNRS.
mining at Grimes Graves, Norfolk, Preliminary Report.
Proceedings of the Prehistoric Society of East Anglia 4 (1), Briart, A., Cornet, F.-L., Houzeau de Lehaie, A. 1868.
113-125. Rapport sur les dcouvertes gologiques et archologiques
faites Spiennes en 1867. Mmoires et publications de la
Armstrong, A.L.1927. The Grimes Graves problem in the Socit des Sciences, des Arts et des Lettres du Hainaut,
light of recent researches. Proceedings of the Prehistoric Troisime srie, tome 2, anne 1866-1867, 355-398.
Society of East Anglia 5(2), 91-136.
Cartailhac, E. 1896. La France prhistorique daprs les
Armstrong, A.L.1934. The Grimes Graves, Norfolk. Re- spultures et les monuments. 2nd ed,vol. I. Paris: Librairie
port on the Excavation of Pit 12. Proceedings of the Pre- Flix Alcan.
historic Society of East Anglia 7(3), 382-394.
Cauwe, N., Vander Linden, M. and Vanmontfort, B. 2001.
Augereau, A.1995. Les ateliers de fabrication de haches de The Middle and Late Neolithic. In N. Cauwe, A. Hauzeur
la minire du Grand Bois Marot Villemaur-Sur-Vanne and P.-L.van Berg (eds.), Prehistory in Belgium. Special
(Aube). In J. Plegrin and A. Richard (eds.), Les mines de issue on the occasion of the XIVth Congress of the Inter-
silex au Nolithique en Europe : avances rcentes, Actes national Union for Prehistoric and Protohistoric Sciences.
de la table-ronde internationale de Vesoul, 18-19 octobre Anthropologica et Praehistorica 112, 77-89, 156-163.
1991, 145-158.
Cels, A. and De Pauw, L. 1885-1886. Considrations sur la
Bonenfant, P.-P. 1993. Spiennes (Mons, Hainaut): mines taille du silex telle quelle tait pratique Spiennes lge
nolithiques de silex. Annales dHistoire de lArt et de la pierre polie. Bulletin de la Socit dAnthropologie
dArchologie de lUniversit Libre de Bruxelles 15, 123- de Bruxelles 4, 245-246.
125.
Centre archologique de Recherche minire (ed.) 1990.
Bonenfant, P.-P. and Auzou, G. 1993. Mons/Spiennes Etude pralable la sauvegarde et la mise en valeur du
Camp--Cayaux, fouilles de minires nolithiques. site archologique de Spiennes - Mons, Spiennes.
Campagne de 1992. Chronique de lArchologie en Wal-
lonie 1, 22-23. Childe, V.G. 1925. The Dawn of European Civilization.
1st ed. London: Routledge, and Kegan Paul Ltd.
Borkowski, W., Migal, W., Saaciski, S. and Zalewski,
M. 1991. Possibilities of investigating Neolithic flint Childe, V.G. 1957. The Dawn of European Civilization.
economies, as exemplified by the banded flint economy. 6th ed. London: Routledge, and Kegan Paul Ltd.
Antiquity 65 (248), 607-627.
Clark, J.G.D. 1952. Prehistoric Europe. The Economic
Bostyn, F. and Lanchon, Y. (eds.) 1992. Jablines: Le Haut Basis. London: Methuen, and Co. Ltd.
Chteau (Seine-et-Marne). Une minire de silex au No-
lithique. Paris: Editions de la Maison des Sciences de Clark, G. 1977. World prehistory in new perspective. Cam-
lHomme, DAF 35, 246 pages. bridge: Cambridge University Press.
72
Clark, G. and Piggott, S. 1933. The Age of the British Collet, H. and Van Neer, W. 2002. Stratigraphie et faune
Flint Mines. Antiquity 7, 166-183. dun puits dextraction nolithique Petit-Spiennes. An-
thropologica et Prehistorica 113, 73-104.
Clarke, W.G. (ed.) 1915. Report on the Excavations at
Grimes Graves, Weeting, Norfolk, March-May, 1914. Collet, H. and Woodbury, M. 2000. Mons/Spiennes: fouille
London: Prehistoric Society of East Anglia. de puits dextraction de silex Petit-Spiennes, Chronique
de lArchologie wallonne 8, 30-31.
Collet, H. 2000. Fouille de nouveaux puits dextraction
Petit-Spiennes et dcouverte de fragments de polissoir. Collet, H. and Woodbury, M. 2001. Mons/Spiennes: pour-
Notae Praehistoricae 20, 163-170. suite de la fouille de puits dextraction de silex au lieu-dit
Petit-Spiennes. Chronique de lArchologie wallonne 9,
Collet, H. 2003a. Mons/Spiennes: indices dexcavations 39-40.
souterraines dans les parcelles 406c et 407a de Petit-
Spiennes. Chronique de lArchologie wallonne 11, 40- Collet, H. and Woodbury, M. 2002. Mons/Spiennes: campagne
42. de fouille de puits dextraction de silex au lieu-dit Petit-
Spiennes. Chronique de lArchologie wallonne 10, 37-38.
Collet, H. 2003b. Mons/Spiennes: fouille dans la parcelle
393c de Petit-Spiennes. Chronique de lArchologie Collet, H. and Woodbury, M. 2006. Mons/Spiennes: pour-
wallonne 11, 39-40. suite des fouilles dans la parcelle 393c de Petit-Spiennes.
Chronique de lArchologie wallonne 13, 47-48.
Collet, H. 2004. Les mines nolithiques de Spiennes: tat
des connaissances et perspectives de recherche. In Sec- Collet, H. and Woodbury, M. in press a. Mons/Spiennes:
tion 9. The Neolithic in Near East and Europe. Acts of the dcouverte dateliers et de puits dextraction du silex rue
XIVth UISPP Congress, University of Lige, Belgium, 2-8 dHarmignies. Chronique de lArchologie wallonne 14.
September 2001, BAR International Series 1303, Oxford:
BAR, 129-133. Collet, H. and Woodbury, M. in press b. Mons/Spiennes: in-
dices miniers nolithiques dans la parcelle 169a au Camp-
Collet, H., Deramaix, I., Sartieaux, P. and Vander Linden, -Cayaux. Chronique de lArchologie wallonne 14.
M. 1997. Fouille prventive de puits dextraction de silex
Petit-Spiennes (Hainaut). Notae Praehistoricae 17, 203- Collet, H. and Woodbury, M. in press c. Mons/Spiennes:
212. dcouverte de vestiges dpoque indtermine rue du
Moulin. Chronique de lArchologie wallonne 14.
Collet, H. and Toussaint, M. 1998. Dcouverte dun
squelette humain nolithique sur le site minier de Petit- Colman, P. 1957. Le Nolithique et ses prolongements
Spiennes (Hainaut, Belgique): tude prliminaire. Interno Spiennes. Bulletin de la Socit royale belge dEtudes
2. Journe dinformation du 14 novembre 1998, Paris, 113- gologiques et archologiques Les Chercheurs de la Wal-
124. lonie 16, 226-290.
Collet, H. and Vander Linden, M. 1998. Petit-Spiennes: Cupillard, Ch., Affolter, J. with Campy M., Contini, D. and
compte-rendu de la campagne 1998. Notae Praehistoricae Richard, H. 1995. La minire de silex nolithique de Blanc-
18, 167-173. Saule Etrelles-et-la-Montbleuse (70) et lexploitation du
silex lacustre oligocne infrieur de Haute-Sane durant
Collet, H. and Vankerkhoven, A. 2004. Mons/Spiennes: le Nolithique, in: J. Pelegrin and A. Richard (eds.), Les
poursuite des fouilles dans la parcelle 393c de Petit- mines de silex au Nolithique en Europe : avances r-
Spiennes. Chronique de lArchologie wallonne 12, 39- centes. Actes de la table-ronde internationale de Vesoul,
40. 18-19 octobre 1991, 179-240.
73
Dchelette, J. 1908. Manuel dArchologie prhistorique, Fechner, K., Gillet, E., Henton, A., Laurent, C., Mathieu,
celtique et gallo-romaine I, Paris. S. and Van Assche, M. 1993. tude archologique du trac
du gazoduc Flobecq-Quvy, Ministre de la Rgion wal-
Defgne, A. and Collet, H. 2003. Mons/Spiennes: premiers lonne, Convention n 9311610, unpublished report.
rsultats palynologiques dans la zone minire de Petit-Spi-
ennes. Chronique de lArchologie wallonne 11, 43-44. Felder, P.J., Rademakers, P.C.M. and De Grooth, M.E.Th.
(eds.) 1998. Excavations of Prehistoric Flint Mines at Ri-
De Koninck, M. 1860. Sur quelques ossements humains jckholt-St. Geertruid (Limburg, The Netherlands) by the
fossiles et sur quelques silex taills. Notice de M. C. Ma- Prehistoric Flint Mines Working Group of the Dutch Ge-
laise. Rapport de M. De Koninck. Bulletin de lAcadmie ological Society, Limburg Section, Bonn: Deutschen Ges-
royale des Sciences, des Lettres et des Beaux-arts de ellschaft fr Ur- und Frhgeschichte e.V. in Kommission
Belgique 10 (29me anne, 2me srie), 511-514. bei Dr. Rudolf Habelt GmbH, (Archologische Berichte
12).
De Laet, S.J. 1982. La Belgique davant les Romains. Wet-
teren: Universa. Gautier, A. and Biondi, S. 1993. Forgotten Faunal Collec-
tion from the Prehistoric Flint Mines at Spiennes (Prov-
Delvaux, E. 1885-1886. Excursion de la socit Mesvin, ince of Hainaut, Belgium). Helinium 33 (1), 80-98.
Spiennes et Harmignies le 5 septembre 1885. Bulletin
de la Socit dAnthropologie de Bruxelles 4, 176-208. Gosselin, F. 1986. Un site dexploitation du silex Spiennes
(Hainaut), au lieu-dit Petit-Spiennes. Vie archologique
De Pauw, L. and van Overloop, E. 1889-1890. Les 22, 33-160.
ateliers prhistoriques de Spiennes. Bulletin de la Socit
dAnthropologie de Bruxelles 8, 28-39. Grooth, M.E.Th. de 1991. Socio-economic aspects of
Neolithic mining: a preliminary study. Helinium, 31(2),
Desterbecq, D. 2004. Mons/Spiennes: fouilles de minires 153-189.
nolithiques dans la parcelle 406b de Petit-Spiennes.
Chronique de lArchologie wallonne 12, 38-39. Grooth, M.E.Th. de 1997. The social context of Neolithic
flint mining in Europe. In R. Schild and Z.Sulgostowska
Desterbecq, D. and Joris, J.-P. 2006. Mons/Spiennes: (eds.), Man and Flint, Warszawa: Institute of Archaeology
fouille de minires nolithiques Petit-Spiennes, Chro- and Ethnology, Polish Academy of Sciences, 71-75.
nique de lArchologie wallonne, 13, 44-45.
Hamal-Nandrin, J. and Servais, J. 1925. Compte-rendu de
Di Lernia, S., Fiorentino, G., Galiberti, A. and Basili, R. fouilles entreprises dans des emplacements dhabitations
1995. The Early Neolithic mine of Defensola A (I 18): et dateliers nolithiques Spiennes (prs de Mons, prov-
flint exploitation in the Gargano area. Archaeologia Polo- ince de Hainaut, Belgique). Bulletin de la Socit prhis-
na 33, 119-132. torique franaise, 22, 73-103.
Dzieduszycka-Machnikowa, A. and Lech, J. 1976. Ne- Hawkes, J. 1965. History of Mankind: cultural and scien-
olityczne zespoy pracowniane z kopalni krzemienia w tific development. Volume One, Part I: Prehistory. New
Sspowie. (Polskie badania archeologiczne 19). Wrocaw- York: The New American Library, Inc.
Warszawa-Krakw-Gdask: Zakad Narodowy imienia
Ossoliskich. Heim, J. 1971. Datation par la palynologie dun fond de
foss de la civilisation Michelsberg Petit-Spiennes. In F.
Evans, J. 1872. The Ancient Stone Implements, Weapons Hubert, Fosss nolithiques Spiennes. Premier rapport.
and Ornaments, of Great Britain. London: Longmans, (Archaeologia Belgica 136), Bruxelles: Service national
Green, Reader, and Dyer. des Fouilles, 62-66.
74
Heim, J., Collet, H., Woodbury, M., Vankerkhoven, A. Hubert, F. and Soumoy, M. 1993. Mons/Spiennes: fouille
and Jadin, I. 2003. Etude palynologique de structures node prvention avant lamnagement de la Station de Re-
lithiques fouilles Petit-Spiennes en 1987, Notae Prae- cherches, Chronique de lArchologie en Wallonie 1, 22.
historicae 23, 141-145.
Jadewski, K. 1984. Urgeschichte Mitteleuropas,
Heinzelin, J. de, Orban, R., Roels, D. and Hurt, V. 1993. Wrocaw: Zakad Narodowy imienia Ossoliskich.
Ossements humains dits nolithiques de la rgion de Mons
(Belgique), une valuation. Bulletin de lInstitut royal des Jeudy, F., Jeunesse, Ch., Monnier, J.-L., Pelegrin, J., Petre-
Sciences naturelles de Belgique, Sciences de la Terre 63, quin, A.-M., Petrequin, P. and Praud, I. 1995. Les carrires
311-336. nolithiques de Plancher-Les-Mines (Haute-Sane). Ex-
emples dune approche intgre. In J. Pelegrin and A. Ri-
Hensel, W. and Tabaczyski, S. 1978. Rewolucja neoli- chard (eds.), Les mines de silex au Nolithique en Europe:
tyczna i jej znaczenie dla rozwoju kultury europejskiej. avances rcentes. Actes de la table-ronde internationale
Wrocaw: Zakad Narodowy imienia Ossoliskich. de Vesoul, 18-19 octobre 1991, 241-280.
Hubert, F. 1969. Fouilles au site minier nolithique de Spi- Krukowski, S. 1939. Krzemionki Opatowskie. Warszawa:
ennes. Campagne de 1965. (Archaeologia Belgica 111). Muzeum Techniki i Przemysu.
Bruxelles: Service national des Fouilles.
Lech, J. 1981. Flint mining among the early farming com-
Hubert, F. 1971. Fosss nolithiques Spiennes. Premier munities of central Europe, Przegld Archeologiczny,
rapport. Annexe de J. Heim, (Archaeologia Belgica 136). (Wrocaw) 28, 5-55.
Bruxelles: Service national des Fouilles.
Lech, J. 1982-1983. Flint work of the early farmers. Pro-
Hubert, F. 1976a. Puits de mines la tranche du chemin duction trends in central European chipping industries
de fer Spiennes. Archaeologia Belgica 186, Conspectus from 4500-1200 BC. An outline. Acta Archaeologica Car-
MCMLXXV, 9-11. pathica (Krakw) 22, 5-63.
Hubert, F. 1976b. Une nouvelle zone minire Spiennes. Lech, J. 1983. Flint mining among the early farming com-
Archaeologia Belgica 186, Conspectus MCMLXXV, 12- munities of central Europe. Part II - Basis of research into
15. flint workshops. Przegld Archeologiczny (Wrocaw) 30,
47-80.
Hubert, F. 1976c. Atelier de taille du silex de lge du
Bronze Spiennes. Archaeologia Belgica 186, Conspectus Lech, J. 1991. The Neolithic-Eneolithic transition in pre-
MCMLXXV, 16-20. historic mining and siliceous rock distribution. In J. Li-
chardus (ed.), Die Kupferzeit als historische Epoche, Part
Hubert, F. 1978. Une minire nolithique silex au Camp- 2, Bonn: Dr. Rudolf Habelt GMbH, 557-574.
-Cayaux de Spiennes. (Archaeologia Belgica 210),
Bruxelles: Service national des Fouilles. Lech, J. 1997. Remarks on Prehistoric Flint Mining and
Flint Supply in European Archaeology. In A. Ramos-Mil-
Hubert, F. 1980. Zum Silexbergbau von Spiennes (B 1). In ln and M. A. Bustillo (eds.), Siliceous Rocks and Culture,
G. Weisgerber, R. Slotta, and J. Weiner (eds.), 5000 Jahre Granada: Editorial Universidad de Granada, 611-637.
Feuersteinbergbau. Die Suche nach dem Stahl der Steinzeit,
Bochum: Deutschen Bergbau-Museum, 124-139. Lech, J. 2004. O badaniach prehistorycznego grnictwa
krzemienia i kopalni w Krzemionkach Opatowskich.
Hubert, F. 1983. Fouille au puits n2 du Camp--Cayaux Przegld Archeologiczny (Wrocaw) 52, 15-88.
de Spiennes (Hainaut), Archaeologia Belgica 253, 17-21.
75
Lech, J. and Leligdowicz, A.1998. Der Silexbergbau. In Lo, A. de and Rahir, E. 1929. Notice sur les fouilles
J. Preu (ed.), Das Neolithikum in Mitteleuropa: Kulturen excutes Spiennes en 1925 et en 1928. Bulletin de la
Wirtschaft Umwelt vom 6. bis 3 Jahrtausend v.u.Z., Socit dAnthropologie de Bruxelles 44, 52-69.
Weissbach, Beier and Beran (Archologische Fachliter-
atur, vol. 1/1), 121-136. Longworth, I., Herne, A. Varndell, G. and Needham, S.
1991. Excavations at Grimes Graves, Norfolk, 1972-1976.
Lech, J. and Longworth, I. 2000. Kopalnia krzemienia Fascicule 3. Shaft X: Bronze Age Flint, Chalk and Metal
Grimes Graves w wietle nowych bada. Przegld Arche- Working. London: British Museum Press.
ologiczny (Wrocaw) 48, 19-73.
Longworth, I. and Varndell, G. 1996. Excavations at
Lech, J. and Longworth, I. 2006. The Grimes Graves flint Grimes Graves, Norfolk, 1972-1976. Fascicule 5. Mining
mine site in the light of two Late Neolithic workshop as- in the Deeper Mines, London: British Museum Press.
semblages: a second approach. In G. Krlin and G. Weis-
gerber (eds.), Stone Age - Mining Age, Der Anschnitt 19, Marin, M.-E. 1961. La priode de La Tne en Belgique.
413-422. Le groupe de la Haine. (Monographies dArchologie
Nationale 2), 92-132. Bruxelles: Muses royaux dArt et
Lefort, M. and Cody, G. 1953. Les cahiers de Spiennes, dHistoire.
Octobre - Dcembre 1953, 2nd ed. fasc. 1.
Mercer, R.J. 1981. Grimes Graves, Norfolk. Excavations
Lefort, M. 1954. Les cahiers de Spiennes. Spiennes. 1971-72: Volume I and II. London: Her Majestys Station-
Fvrier - mars 1954, fasc. 2. ery Office.
Lichardus, J., Lichardus-Itten, M., Bailloud, G., Cauvin, J. Milisauskas, S. and Kruk, J. 2002. Middle Neolithic:
et al. 1985. La Protohistoire de lEurope. Le Nolithique et continuity, diversity, innovations, and greater complexity,
le Chalcolithique entre la Mditerrane et la mer Baltique. 5500/5000-3500/3000 BC. In S. Milisauskas (ed.), Euro-
Paris: Presses Universitaires de France. pean Prehistory. A Survey, 193-246. New York: Kluwer
Academic/Plenum Publishers.
Lo, A. de 1913. Les fouilles de M. Louis Cavens
Spiennes en 1912. Bulletin des Muses royaux du Cin- Mortillet, G. de 1885. Le prhistorique, antiquit de
quantenaire 12, 36-37 and 44-46. lhomme. 2nd ed. Paris: C. Reinwald.
Lo, A. de 1914. Les fouilles de M. Louis Cavens Pelegrin, J. 1995. Rflexions mthodologiques sur ltude
Spiennes en 1913. Bulletin des Muses royaux du Cin- de sries lithiques en contexte datelier ou de mine. In J.
quantenaire 13, 35-37. Pelegrin and A. Richard (eds.), Les mines de silex au No-
lithique en Europe: avances rcentes. Actes de la table-
Lo, A. de 1925. Notice sur les fouilles excutes ronde internationale de Vesoul, 18-19 octobre 1991, 159-
Spiennes en 1912, 1913 et 1914. Bulletin de la Socit 165.
dAnthropologie de Bruxelles 40, 151-171.
Piggott, S. 1954. The Neolithic Cultures of the British
Lo, A. de and Munck, E. de 1891. Notice sur des fouilles Isles, Cambridge: Cambridge University Press.
pratiques rcemment sur lemplacement du vaste atelier
nolithique de Spiennes (Hainaut). Section: Ateliers et puits Robaszynski, F. and Christensen, W.K. 1989. The upper
dextraction de silex en Belgique, en France, en Portugal, Campanian-Lower Maastrichtian chalks of the Mons ba-
en Amrique. In Congrs international dAnthropologie sin, Belgium: a preliminary study of belemnites and fo-
et dArchologie prhistoriques. Compte-rendu de la dix- raminifera in the Harmignies and Ciply areas. Geologie en
ime session (Paris: 1889), Paris, 569-602. Mijnbouw 68, 391-408.
76
Robert, C. 2003. Mons/Spiennes: les minires nolithiques Verheyleweghen, J. 1953. Dcouverte sur le territoire
de Petit-Spiennes. Chronique de lArchologie wallonne de Spiennes dune phase doccupation des hommes de
11, 42-43. Michelsberg, antrieure celle du plateau du Camp--
Cayaux (Compte-rendu des fouilles entreprises en 1953).
Saville, A. 1981. The Flint Assemblage. In R. J. Mercer, Bulletin de la Socit royale belge dAnthropologie et de
Grimes Graves, Norfolk. Excavations 1971-7, Volume II, Prhistoire 64, 141-162.
1-178. London: Her Majestys Stationery Office.
Verheyleweghen, J. 1962. Un dpt funraire de crne
Scollar, I. 1955. A Neolithic Enclosure at Spiennes. nolithique Spiennes (Hainaut). Avec une note concern-
Antiquity 29 (115), 159-161. ant limportance de la population spiennienne et un essai
sur la structure sociale de la tribu. Helinium 2 (3), 193-
Tabaczyski, S. 1970. Neolit rodkowoeuropejski: pod- 214.
stawy gospodarcze. Wrocaw: Zakad Narodowy imienia
Ossoliskich. Verheyleweghen, J. 1963. Evolution chronologique du
Nolithique au Camp--Cayaux de Spiennes (Hainaut)
Tabaczyski, S. 1972. Gesellschaftsordnung und Gter- avec une tude statistique dimensionnelle et caractrielle
austausch im Neolithikum Mitteleuropas. In B. Thaler du matriel lithique. Helinium 3 (1), 3-38.
(ed.), Neolithische Studien I, Berlin: Akademie-Verlag,
(Wissenschaftliche Beitrge der Martin-Luther-Universitt Verheyleweghen, J. 1966. Le Nolithique minier belge,
Halle Wittenberg 1972/1, L 7), 31-96. son origine et ses relations culturelles. Second Colloque.
Palaeohistoria 12, Acta et Communicationes Instituti Bio-
Toilliez, D. 1851. Troisime notice sur des antiquits d- Archaeologici universitatis Groningae. Proceedings of the
couvertes dans le Hainaut. Bulletin de lAcadmie royale Second Atlantic Colloquium (Groningen 6-11 April 1964),
de Belgique 18 (1re partie), 659-670. 529-557.
Toussaint, M., Collet, H. and Vander Linden, M. 1997. Vermeersch, P.M., Vynckier, G. and Walter, R. 1990.
Dcouverte dun squelette humain dans le puits de mine Thieusies Ferme de lHost, site Michelsberg. II Le
nolithique ST 11 de Petit-Spiennes (Hainaut). Notae matriel lithique. (Studia Praehistorica Belgica 6). Leu-
Praehistoricae 17, 213-219. ven: KULeuven.
Vandevelde, P. and Hubert, F.1987. Deux vases Michelsberg Weisgerber, G., Slotta, R. and Weiner, J. (eds.) 1980. 5000
de Spiennes et le matriel associ. Bulletin de la Socit roy- Jahre Feuersteinbergbau. Die Suche nach dem Stahl der
ale belge dAnthropologie et de Prhistoire 98, 223-234. Steinzeit. Bochum: Deutschen Bergbau-Museum.
Vanmontfort, B. 2004. Converging Worlds. The Neolithisa- Whittle, A. 1996. Europe in the Neolithic. The creation of
tion of the Scheldt basin during the late fifth and early fourth new worlds. Cambridge: Cambridge University Press.
millennium calBC. Unpublished PhD. Leuven: Katolieke
Universiteit Leuven, Department Archeologie, Kunstweten- Zawisza, J. 1873. Kongres Antropologii i Archeologii
schappen en Musicologie, Afdeling Archeologie. Przedhistorycznej, w Bruxelli 1872 r. Warszawa.
Vanmontfort, B., Collet, H. and Cromb, Ph. (in press). Les Acknowledgments
Industries lithiques tailles des 4me et 3me millnaires
dans les bassins de lEscaut et de la Meuse (Belgique). In The study of the material from the Stevens collection kept
Actes du colloque Les industries lithiques tailles des 4me at the Royal Belgian Institute of Natural Sciences was
et 3me millnaires en Europe occidentale, Toulouse, 6-9 granted by the Synthesys BE-TAF programme under the
avril 2005. reference 1055.
77
78
Franoise Bostyn, Franois Giligny, Adrienne Lo Carmine: A New Flint Mine at Flins-sur-Seine/ Aubergenville (Yvelines, France)
A New Flint Mine at Flins-sur-Seine/ Aubergenville

(Yvelines, France)
Franoise Bostyn, Franois Giligny, Adrienne Lo Carmine
Abstract: Discovered in the 1920s, and located in the Seine valley downstream from Paris, the mine of Flins-sur-Seine/
Aubergenville was identified as a flint mine from recent aerial photographs. The site of Flins-sur-Seine exploited Bar-
tonian Tertiary flint to manufacture axes. Surface collection and geophysical surveys, underway since 2002, show that
the area with knapping waste covers about 15 hectares and that the extraction zone extends over at least 3 to 5 hectares.
Evidence of first technological stages of axe manufacture, such as bifacial preparation, roughouts and knapped axes/axe-
cores are present, but neither polished artefacts nor traces of settlement were found.
Keywords: Flins, flint mine, Bartonian flint, geophysical surveying, axe production.
1. Localisation of the mine and research

history
The mine of Flins-sur-Seine / Aubergenville, to the West Clos (De Sacy and Baudouin, 1926 and 1927). The varie-
of Paris, dominates the Seine valley on the right bank of ty of axes and the small number of polished axes prompted
one of its tributaries, the Mauldre. The mine sits atop a them to attribute this material to the end of the Campig-
plateau at an altitude of between 100 and 125 m, along a nien (terminology of the time). These authors also noted
slight southeast / northwest slope (Fig. 1a). In this area, the resemblance of the material at Flins with that of the
the Seine valley, as well as the secondary valleys of the Jablines, classifying Flins-sur-Seine in the upper Campig-
Mauldre, and Vaucouleurs, cut deeply into the Tertiary nien, or Jablinien (now obsolete terminology).
geological substratum (Eocene, Oligocene) as far as the
secondary formations (Cretaceous). White chalk with The site has been the object of several surface surveys,
flint nodules lies at the base of the stratigraphic sequence, some of which have been published or officially reported
which is covered over by Ypresian and Lutetian layers. (Daniel, 1965; Bourgeois, 1982). The mine has been regu-
Limestone and marl of the Upper Bartonien (Ludien) or larly surveyed since the 1980s by many enthusiasts who
Lower (Saint-Ouen limestone) complete the stratigraphy. have, over the years, made up impressive collections both
Siliceous material is abundant and varied, and the Saint- in quantity and quality, but the site has never been exca-
Ouen limestone is known for its brown to light beige flint vated. As part of a collective research programme begun
found in slabs or nodules (Fig. 1b). in 2000, we have undertaken a synthesis of this wealth of
One of the earliest references to the mine of Flins-sur- data by through two complementary approaches: mapping
Seine dates from 1926, with the work of S. de Sacy and the site by using different surveying methods, and re-ex-
M. Baudouin concerning the location, characterization and amining the artefacts and the historical data from the per-
cultural attribution of items from the small hamlet of Le spective of the technologies of production.
79
ine
25 m
La Se
50 m
La M
100 m
auld
r
Flins-sur-Seine/
e
Aubergenville
Epne 175 m
La R
ou
ase
Rau
mine shafts
Jumeauville waste
rough-outs & waste
settlement area
ic e
h
R
de
Ru
0 5 km
Fig. 1: A. Location of the mine Flins-sur-Seine/Aubergenville. K. Larsen del. 1999
La Seine
limon
Stampien : Fontainebleau sand
Stampien inf. : Ludien
Bartonien inf. : Saint-Ouen limestone
alluvium
Santonien-Coniacien-Campanien 1 km.
Fig. 1: B. Geological map (BRGM).
80
2. Recent field data (2000-2003) n=

50
Since 2000, different surveying methods have been gradu- 45 south-east conc.
40 east concentration
ally brought into play. Their continuous interaction has so
35
far produced relevant results in different sectors of the site.
30
The objectives of the overall survey are to establish the
25
limits of the mining areas in relation to the size of the flint
20
products.
15
10
2.1. Aerial surveying 5
0
In order to make an archaeological map, aerial photographs 1 2 3 4 5 6 7 8 9 10 11 m
of the mine were taken by the Archaeological Service of the
Fig. 3: Distribution of the distances between the anomalies identified on
Department of Les Yvelines in 1999. Adjusted and placed aerial photographs (pit) for two concentrations (metres).
on cadastral surveys of the area, the photographs allowed
us to identify the presence of mine shafts in certain areas first of 0.37 ha and the second of 1.5 ha, 96 anomalies
(work done by Terra NovA: 2001). Three concentrations were recorded. Distances between the structures are slight,
appear from east to the west. Based on anomalies visible in most cases less than 5 m. The average distance between
in aerial photography, interpreted as mining structures, an the structures in the Eastern zone is 4.3 m, while this av-
estimate of the number of such structures was made in the erage is 3.5 m in the central area. Comparing distribu-
areas of interest. Two test areas were chosen for further in- tion histograms of the distances for the two areas shows a
vestigation: one in the central part, the second in the east- greater coefficient of variation in the Eastern zone, struc-
ern part of the site. turally less dense than the other (Fig. 3). Almost one third
of the distances in the Central zone are smaller than 3 m,
The central point of the anomalies was pinpointed and its compared to one fifth for those in the Eastern zone. How-
coordinates recorded in a GIS in order to make an accurate ever, the Eastern zone presents a fairly large number of
count of the anomalies to calculate the average distance distances greater than 5 m (29%), a value which is signifi-
between the structures (Fig. 2). Within the two areas, the cantly smaller in the other sector (7%).
Fig. 2: Location and density of mine shafts interpreted according to aerial photographs in the northern sector of the site (Photography: P. Laforest Service
Archologique Dpartemental des Yvelines, adjustment: Terra NovA).
81
mine shafts
Fig. 4: Electrical resistivity map at 1 m treated with a high-pass filter in the middle part of the site (Terra NovA document, 2003, Riquier C. DAO).
These distances suggest the presence of shallow extraction 2.2. Geophysical surveying
pits. By comparing these data with the map of the Jablines
surface mine, we can envisage pits at a depth of between Two geophysical surveying methods were tested on a sur-
2 and 4 m for those which are closest together, as in the face where aerial photography had suggested mine shafts:
North zone at Jablines, where the land is stripped and one using a magnetometer and one using electrical methods
where structures are the closest together and the most shal- (Aubry et al. 2002, 2003). After this first experiment, the
low (Bostyn and Lanchon eds 1992). Without prejudging magnetic method was found to be inappropriate because it
the depth of mining structures at Flins since they have not could not detect the mine shafts; the perturbations recorded
been excavated, the 2003 experiment identified a layer of by this method were caused by changes in the soil composi-
flint at an approximate depth of 2.50 m. The convergence tion and agricultural activities. Yet, on the basis of a meas-
of these various observations suggests that some of the ex- urement for every metre, the electrical survey confirmed the
traction sites were dug to a shallow depth. presence of structures visible in aerial photography, with a
good resolution.
The spatial extent of the mine could not, however, be de-
termined with reliable precision, as most of the field sur- Having thus been proven effective, this method was applied
faces photographed revealed little because of a lack of to other plots, in particular those for which the aerial pho-
vegetation. Thus, geophysical and pedestrian surveys were tography had revealed nothing. We applied the electrical
conducted concurrently. method to cover a total surface area of more than 5 hectares
82
Fig. 5a: Mapping the number

of chips per square.
Fig. 5b: Mapping the number of roughouts

per square (2002-2003 survey, Riquier C.
DAO).
(Fig. 4). The interpretation of the resistivity maps allowed fix the limits of the site approximately. The relative impor-
us to target the southernmost point of the research area as tance of material recovered allowed us to define areas of
the development zone for extraction structures, while the concentration, corresponding to axe blade knapping activi-
northernmost sectors do not seem to have not been used ties. The artefacts found were mostly chips, tools of poor
for underground extraction. However, periglacial phenom- craftsmanship and fragments of bifacial pieces. And ab-
ena (frost polygons) are clearly visible. sence of ceramics and grinding equipment correlated with
the small number of tools suggests that domestic activity
2.3 Pedestrian surveys on the site was uncommon. The area was used as a mine
and workshop. A few test boulders and some large sand-
Systematic pedestrian surveys were carried out in two suc- stone hammerstones were also collected.
cessive campaigns on two different plots covering 3 hec-
tares. The Northwest Sector was prospected in 2002, the The distribution map of the number of chips per square
Southeast Sector in 2003. (Fig. 5a), presents a greater concentration in the Southeast
Sector, within a strip 50 meters wide, situated northeast /
The sampling was at a 25% rate for 100 m2 squares. Fur- southwest, that follows the level curve of 110 m NGF. This
thermore, a quick walk over the whole area allowed us to sector has the highest densities (> 120 chips/ square). Two
83
0 250 m
mine shafts (aerial photograph) pedetrian prospection 2002-2003
probable shaft extent extent of the workshops according to flint waste
geophysical prospection 2001-2003 experimental shaft excavation
Fig. 6: Extent of the site and summary of the surveys.
areas of lower density appear in this sector at the southeast (Fig. 6) is only possible for the Southeast Sector. In this
and northwest boundaries of the plot, at the highest and case, there is a strong correlation between the areas of
lowest points. The density is much lower (to 90 chips / debitage waste and the anomalies, which are concentrated
square) in the Northwest Sector. An area of concentration in the southern half of the plot. This is also, topographi-
emerges in the Northwest, along the path to the site. The cally speaking, the highest zone and the one most subject
distribution map of all rough pieces offers the same image to erosion. In the Northwest Sector, where only pedestrian
(Fig. 6). However, the distribution map of the roughouts surveys and geophysical data were possible, no anomaly
(Fig. 5b) gives a contrasting image, with a higher density was detectable by geophysical surveying, while relatively
in the Northwest Sector. In this sector, the concentration high concentrations of debitage waste, tools and roughouts
of roughouts coincides with that of the chips, while in the were confirmed. This sector is thus representative of a zone
other sector, the distribution is more diffuse across the en- restricted to knapping, apart from the areas of extraction.
tirety of the surface and does not outline any concentration
different from the chips. 3. Production
2.4. Conclusion The analysis of the waste flint and the tools found during
surface collection showed that the knapping activities at
A comparison between the data offered by the pedestrian the mine were primarily related to the production of axes.
surveys, the geophysical data and the aerial photography Very few other tools usually associated with settlement
84
(scrapers, blades) were found. The flint artefacts found at 3/4 of the piece - or crushed with a hammerstone.
the mine may be divided into several categories represent-
ing various stages of the chane opratoire of axe produc- The angles of the edges are considered irregular when they
tion, but also of other types of products: vary at least 30 from the edge. Indeed, some axe blades
- rough boulders and test boulders have angles on part of the edge at more than 90 degrees, a
- debitage waste including chips, blades, and splinters result considered as a technical error.
- bifacial fragments in various manufacture stages, or oth-
er flint pieces deriving from this chane opratoire, from The bifacial pieces from the surface collections were stud-
the removal platelet down to the retouched axe ied in greater detail. Within a total of 542 pieces, only those
- nuclei and other products not connected with the chane found completely intact were retained, reducing the total
opratoire of axe manufacture sample to 428. The blanks are platelets, plate fragments,
- debris chips, chips due to frost, or classed as uncertain. The main
- flake tools, blades or debris blanks used are mostly raw plates and chips, in equal pro-
portion (Fig. 7). A 6% ratio of frost chips, identified solely
Since the main purpose for the study of the axe blades is on the steps prior to the knapped axe, should also be noted
to define the causes of abandonment and then to assess the (Fig. 10). Some of the pieces bear ancient fractures related
skill level of the knappers, several criteria were chosen for to the knapping, or to the raw material. However, knappers
systematic recording (in addition to the usual criteria of continued knapping and hewing and the fracturing was not
size, state of preservation, etc.): the quality of the raw ma- the main cause of rejection.
terial, the nature of the blank, and the problems associated
with knapping and related technical errors. The main difficulty of our study is to determine what led
to completion of the pieces and / or the skill level of the
The raw material can be classified as good, moderate or knappers. While it seems obvious for certain criteria, such
poor quality. To define these quality levels, we detailed as errors in size, all other indications are difficult to judge.
the imperfections of the raw material: an intrusive cortex, The presence of cortex and its extent are the criteria gen-
weak silicification, former fracture surface and cavities. erally taken into account to define the degree of progress
A weak blank is defined by a failure to overcome tech- of the pieces, but these can also testify to the expertise of
nical imperfections in the raw material and to obtain an the knapper, since the presence of marginal chips is a tell-
axe blade; this corresponds in our study to the criterion tale sign of a low level of competence. We must therefore
of rejection of pieces in the chane opratoire. A blank of modulate several criteria to determine reasons for aban-
moderate quality may have imperfect silicification and donment, and to attribute the pieces to a particular techni-
thin slab heterogeneity, but still serve as the basis for mak- cal stage.
ing a finished product.
Symmetry in the sections and faces was noted, with a pres-

ence or absence of the axis of symmetry on each piece. The number %
edge, where existing, was described, including whether it
u n ifa c ia l p re p a ra tio n 10 2,3%
was of convex, straight or irregular form, its symmetry and
uni- & bifacial preparation 40 9,3%
its regularisation.
Bifacial unilateral preparation 23 5,4%
Bifacial bilateral preparation 142 33,2%
Technical errors were described for each face, with de-
R o u g h -o u t 144 33,6%
tailed notes regarding hinged flakes, due to the quality of
axe 69 16,1%
the raw material or the knapping, plunging flakes, concave
To ta l 428
flakes, overhewn or minor flakes. The edges are straight,
irregular, or considered regularized when they have been
Fig. 7: Number of pieces at Flins-sur-Seine showing various stages in the
retouched along the entire length of the edge - or at least production of axe manufacturing.
85
10 cm
Fig. 8: 1. Bifacial and bilateral preparation and 2. Bifacial unilateral preparation (Drawing Carmine A. Lo). Flins-sur-Seine.
The technological study helped clarify the various stages of of the process, from unifacial preparation to the finished
the chane opratoire of production of axe blades. The blank product (Fig. 7).
used is mostly of rough slab, or previously struck chips.
The early stages of shaping occur the least frequently in
Despite technical limitations related to the morphology of the sample (17% of the total sample), in particular frag-
the blanks, the chane opratoire vary only slightly. These ments in unifacial unilateral preparation, or bifacial unilat-
sequences are divided into four technical stages: unifa- eral preparation. The more advanced stages, i.e., of bifacial
cial then bifacial preparation, unilateral and bilateral, the and bilateral preparation (Fig. 8) and the waste flakes (Fig.
roughing out, regularization/ retouching and polishing. 9) are the most common (66.8% of the total sample). The
Traces of polishing are not attested at the mine. The bifa- final step, the axe or preform ready to polish, makes up a
cial pieces found are representative of the first three stages little over 16% of the total sample (Fig. 10).
86
2
1
3 4
10 cm
6
5
0
Fig. 9: Roughouts. Flins-sur-Seine (Drawing Carmine A. Lo).
The nodules found are between 8 and 14 cm, though larger with thicknesses of 2.9 cm for the bifacial preparations,
pieces exist as well. Their lengths are between 15 and 25 2.5 cm for the axes.
cm. The pieces obtained from levelling up the stone facing
are larger, especially when the blank is a small plate, as the These flint pieces were discarded mostly because of the poor
rough-hewn large flakes broken off with a hammerstone quality of the raw material, which contains cavities (32%),
greatly reduce the size of the original blank (Fig. 11). The intrusive cortex (33%) or a weak silicification (15%). Only
pieces have an average length of 12.5 cm for those of bifa- 46% of the bifacial and bilateral preparations are made of
cial preparation, 10.5 cm for axes, with a low coefficient of good quality blanks, with 53% for the axes. However, if
variation. These two categories make up the largest and the one takes into account three other criteria, there arises a
smallest pieces (from 6.3 to 26.5 cm). The average widths variety of problems that led to discarding these pieces.
are 5.5 cm for bifacial preparations, 4.2 cm for the axes; Some of them are too hollow on one or both sides: 34%
87
3 4 10 cm
5 6
Fig. 10: Axes. Flins-sur-Seine (Drawing Carmine A. Lo).
of bifacial and bilateral preparations, 37% of the rough- to the statistics: although less frequent, over-levelling was
outs but only 12% for the axes. Another error recognized revealed on 19% of the bifacial bilateral preparations, on
on discarded fragments is repeated blows, relentlessness 13% of the roughouts and on 9% of the axes.
(acharnement) on one or both edges. Sometimes percus-
sion action along the edge caused it to splinter irregularly, An overall analysis of the products found at the mine
creating a defect impossible to correct, as was seen in 61% suggests that the general level of quality was moderate,
of the bifacial preparations, 55% of the roughouts and 40% both in regard to the rough blanks as to the levels of ex-
of the axes. Manual awkwardness is without doubt one of pertise. These results contrast with the level of investment
the most frequently observed problems, impugning the required to acquire the flint, and also with the quality of
poor level of competence of the knapper. To conclude this the products found outside the context of the mine, a point
summary review of inhibitory over-levelling, we will refer to which we shall return later. How should we interpret
88
bifacial bilateral preparation rough-out

30 30
25 25
20 20
length
length
15 15
long
10 10
5 5
0 0
0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14
width width
axe
30
25
20
length
15
10
5
Fig. 11: Length width relationship of bifacial pieces
according to stages in the chane opratoire. 0
0 2 4 6 8 10 12 14
width
35
these findings from a socio-economic viewpoint? It seems

30
that we are faced with two types of productions, with vary-
ing end results. A production of large axe blades (30 cm), axes (Flins)
probably made from platelets selected from the most ho- 25 axes (other sites)
polished axes
mogeneous rock, was carried out at the mine by probably
experienced knappers. These products were intended to
20
circulate at shorter or longer distances and are only rarely
length (cm)
found among the extractive waste. This consideration is

15
logical, since we can expect a very low failure rate on the
part of experienced knappers. With this in mind, we can
consider the presence of apprentice knappers at the loca- 10
tion of the extraction, where the raw material is abundant,
thus allowing for a certain amount of waste. A poor mas-
5
tery of the technical and visualizing know-how (Pelegrin
1991) would result in failed production at different stages
of the chane opratoire. One can also consider that the 0
2 4 6 8
apprentice knappers were intended to produce pieces of
width (cm)
smaller size, the manufacture of which is less risky. The
tools produced may have been used on the spot or taken to
Fig. 12: Comparison of the scale of knapped axes from the mine, with
nearby habitats. knapped and polished axes at a regional level.
89
4. Product diffusion
To assess the extent of product diffusion at the regional which applies to both axes and polished axes (Fig. 12).
level, cartographic work was carried out, mapping the The status of these large artefacts, unfortunately found out
distribution of polished, percussioned and unfinished axe of context, such as in Bonnires-sur-Seine, raises the ques-
blades. This mapping focused primarily on the Depart- tion of their function (Fig. 13, n 1). They do not seem to
ments of Yvelines and Val dOise, but also extended fur- have been used (the sharp edge was retained), and could
ther to the west and southwest. be classified in the category of objects of prestige. In the
context of settlements, the flint artefacts are smaller, often
The flint pieces found outside the mine at a regional lev- recycled, with a reknapped edge; some were used as ham-
el are sometimes larger, 15 cm to 30 cm, an observation merstones, and highly fragmented (Fig. 13, n 2).
1
10 cm
Fig. 13: Polished axes. 1. Discovered out of context. 2. Late Neolithic collective grave
Val de Reuil Les Varennes, 3. Early Neolithic settlement Neauphle-le-Vieux
(1: drawing Lo Carmine A.; 2: Billard C. 1995, fig. 13; 3: drawing Martial E.).
90
Distribution maps of the products show differences de- Polished axes in Bartonian flint are very rare at the sites
pending on the operating stages. The early stages of prepa- of the former Danubian Neolithic (Villeneuve-Saint-Ger-
ration, in particular bilateral and bifacial preparations, are main). For this time frame, polished pieces from the Paris
found almost exclusively at the mining and surrounding Basin are equally rare, and usually made of material other
areas (Fig. 14a). The few points further afield, however, than flint, such as quartzite sandstone or resistant exog-
show that some unfinished pieces were circulated in small enous rock. However, it should be noted that a fragment
quantities. of a polished axe in Bartonian flint was found at the site of
Neauphle, Yvelines (Giligny et al. 1996 and Fig. 13, n3),
Away from the mine, but probably in direct correlation to and an incomplete bifacial piece was found at the site of
it, the highest concentrations correspond to probable flint Ocquerre, Seine-et-Marne (Praud et al. 2002).
knapping workshop sites (Fig. 15). It is useful to recall that
several grooved abraders have been found in the immediate Bartonian flint axes become common during the Middle
area close to another workshop at Jumeauville. During the Neolithic. The Neolithic settlement site at Louviers (Eure),
Neolithic period, the banks of the Mauldre must have been fifty kilometres west of the mine, illustrates the circulation
a particularly dynamic area of activity focusing around of finished products from this mineral source at that time.
the working of flint (Fig. 15). One central site could have At Louviers, tertiary flint makes up 1/3 of all axe blades on
played a role in an interregional exchange system. the site (Giligny ed. 2005).
The distribution of roughouts and axes is more scattered, The deposition of axes in some Neolithic graves indicates
showing a diffusion to the North, on the right bank of the that the production lasted until the late Neolithic, with a
Seine, and about 60 km to the Southwest (Fig. 14b and 16 wide chronological bracket concerning the use and reuse
a). Polished axes were circulated at a much wider distance of collective graves over a long period of time (Seine-
and a much larger scale (Fig. 16b). Oise - Campaniforme Marne). These collective graves are
found in Luzarches Compans Le Val dOise, Presles II La
There is still a difference in quantity between the regions Pierre Plate, LEtang La Ville Cher Arpent (Yam 2005),
south of the Seine and those in the north. The quantity of and Val de Reuil Les Varennes (Billard et al. 1995: Fig.
polished axes in Bartonian flint is particularly low in the 13, n 2).
Val dOise (16%): in this case, the Seine must have been
a natural obstacle (Fig. 17). The presence of workshops Conclusion
to make secondary flint axes in the two regions can also
provide some explanation. In the Yvelines, the two raw These initial analyses show a relatively complex organisa-
materials are represented in the same way, so they must tion in the production of axe blades, in their socio-eco-
have been in direct competition. nomic organisation, where the products seem to have been
made by knappers with highly variable levels of skill, and
To conclude, it should be noted that some of the flint ar- also in the spatial distribution between the areas of extrac-
tefacts travelled at even greater distances, towards Nor- tion strictly speaking the workshop areas, and the polish-
mandy, in the Seine valley and more than 150 km away ing zones. Continuous evidence of tertiary flint axes in
into the plains of Caen and Argentan (Watt 2007, Bostyn both excavation and surface campaigns is a prerequisite
et al. in press b). for a study of their distribution. Such research will better
measure the importance of the mine of Flins at a regional
5. Chrono-cultural context level, which extends more widely into the western Paris
Basin, but also will help evaluate the relative positions of
In the absence of direct excavation-derived data at the Flins the two mines identified with certainty in the Tertiary Ba-
mine, its chronology must be estimated from other exca- sin, that of Flins and that of Jablines.
vated sites. Unfortunately, the longest, most complete prod-
ucts of this mine were found out of context (Fig. 13, n1).
91
Fig. 14a: Distribution map. Bifacial

preparations.
1
2-3
4-5
6-9
> 10
hydrography
department limit
gographical limit of the study
Fig. 14b: Distribution map. Roughouts.
11
22--33
44--55
66--99
>>10
10
hydrography
department limit
1
VAL D'OISE
2
Fig. 15: Map of the different workshop
and polishing sites along the Mauldre
river.
3 4 1- Banthelu
2- Longuesse
5
6
7 3- Guerville
4- Flins
8
PARIS 5- Rosay
9
10 6- Jumeauville
7- Maule
8- Montainville
departments limits
rivers 9- Beynes
YVELINES
tertiary flint 10- Villepreux
secondary flint
axes worshops (tertiary flint)
axes worshops (secondary flint)
????
92
Fig. 16a: Distribution map. Axes.
11
22 -- 33
44 -- 55
66 -- 99
>> 1010
hydrography
department limit
Fig. 16b: Distribution map. Polished axes.
11 - 2
23--35
45--13
5
614- 9- 41
>>1041
hydrography
department limit
others/ undetermined
tertiary flint
secondary flint
4
1 Vexin franais
2
3
Paris
Mantois
6
Fig. 17: Nature of material used for Armorican
polished axes in the north and south of the massif
Seine river. ?
0 100 km limit of tertiary basin
mining complex sites with bartonian axes bartonian axes >1

1 : Bretteville-le-Rabet, 2. Flins-sur-Seine, 3. Jablines, 4. Hardivilliers-Troussencourt, 5. Louviers, 6. Fort-Harrouard
93
Franoise Bostyn Bostyn, F. and Lanchon, Y. (eds) with Boguszewski, A.,

Inrap - UMR 7041 Frugier, C., Jrmie S. et al. 1992. Jablines, Le Haut Ch-
36 alle Thals teau (Seine-et-Marne) : une minire de silex au Nolithi-
59650 Villeneuve dascq que, Documents dArchologie Franaise 35. Paris: Mai-
e-mail: francoise.bostyn@inrap.fr son des Sciences de lHomme.
Franois Giligny Bostyn, F., Cayol, N., Giligny, F. and Lo Carmine, A. in

Universit de Paris I - UMR 7041 press a. Creusement exprimental dun puits dextraction
MAE 21, avenue de luniversit de silex sur la minire de Flins-sur-Seine (Yvelines), Actes
92023 Nanterre du XXe Colloque interrgional sur le Nolithique. Luxem-
e-mail: francois.giligny@univ-paris1.fr bourg : Archologie Luxembourgeoise.
Adrienne Lo Carmine Bostyn, F., Couderc, J., Giligny, F., Lethrosne, H., Le
Universit de Paris I - UMR 7041 Maux, N. and Lo Carmine, A. in press b. Ateliers de fa-
MAE 21, avenue de luniversit brication de haches et minires silex dans lOuest de
92023 Nanterre llede-France (Yvelines, Val dOise). Table ronde de la
e-mail: adrienne_locarmine@yahoo.fr Socit Prhistorique Franaise, mars 2007.
References Bourgeois, L. 1982. Inventaire dune srie lithique pro-

venant de Flins-sur-Seine (Yvelines). Rapport. Versailles:
Aubry, L. and Barrs, E. 2002. Prospections gophysiques, Service Archologique Dpartemental des Yvelines.
Minires nolithiques de Flins-sur-Seine. Campagne
dacquisition. Bricon, D. and Vidal, G. 1983. Haches tailles de grandes
dimensions de la rgion de Flins. Bulletin du Centre de
Aubry, L., Barrs, E. and Marmet, E. 2003. Prospections Recherches Archologiques de la Rgion Mantaise 6, 19-
gophysiques, Minires nolithiques de Flins-sur-Seine. 20.
Synthse des campagnes 2002-2003, Terra NovA.
Daniel, R. 1965. Les ateliers campigniens de Flins (Seine-
Billard, C., Guillon, M., Sunder, F. and Arbogast, R-M.1995. et-Marne). Bulletin de la Socit Prhistorique Franaise
La spulture collective du Val-de-Rueil Les Varennes et les 62, 279-286.
structures associes de lAge du Bronze. In Actes du Collo-
que interrgional sur le Nolithique, Evreux 1993, Revue Giligny, F., Martial, E. and Praud, I. 1996. Le Villeneuve-
Archologique de lOuest, supplement 7, 155-182. Saint-Germain de Neauphle-le-Vieux le Moulin de Let-
tre (Yvelines), Interno 1, 15-32.
Billard, C., Guillon, M. and Verron, G. (eds) with Argo-
bast, R.-M., Cedran, Y., Sunder, F., Dasturgue, J., Gervais, Giligny, F., Martial, E. and Praud, I., with Le Gall, J.
V., Leuge, F., Querre, G. and Salanova, L. Un systme de 1998a. Loccupation des Yvelines au Nolithique ancien:
spultures collectives au Nolithique final Val de Reuil et premiers lments. Rapport dtude. Versailles: Service
Portejoie (Eure), unpublished. Archologique Dpartemental des Yvelines.
Bostyn, F. 2003. De la lame la hache : contextes Giligny, F., Martial, E. and Praud, I., with Bostyn, F., Le
gologiques et socio-conomiques des productions en si- Gall, J., 1998b. Premiers lments sur loccupation des
lex Tertiaire bartonien du Bassin parisien au Nolithique. Yvelines au Nolithique. Interno 2, 43-56.
InLes matires premires lithiques en prhistoire, actes de
la table ronde dAurillac, 20-22 juin 2002, Prhistoire du Giligny, F. and Lo Carmine, A. 2001. Le Nolithique des
Sud-Ouest, supplment 5, 63-70. Yvelines. Rapport dtude 2001, UMR 7041 Nanterre.
94
Versailles : Service Archologique Dpartemental des Yam, S. 2005. Le mobilier funraire des alles couvertes
Yvelines. du Val dOise. Mmoire de Matrise de lUniversit de
Paris 1, 2 vol.
Giligny, F. (ed.), Bostyn, F., Couderc, J., Durand, S., Du-
rand, J., Lo Carmine, A., Martial, Praud, I. and Riquier, C.
2005. Le Nolithique des Yvelines. Rapport intermdiaire
du Projet Collectif de Recherches.
Giligny, F. (ed) 2005 - Louviers la Vilette (Eure) : un

site Nolithique moyen en zone humide. Documents
Archologiques de lOuest, 2005, 343 pages.
Lo Carmine, A. 2000. La production de haches en silex

tertiaire dans le Nord des Yvelines. Mmoire de matrise
de lUniversit de Paris 1.
Lo Carmine, A. 2002. Grandes lames de haches et ateliers

de fabrication dans le nord des Yvelines. Mmoire de DEA
de lUniversit de Paris 1.
Pelegrin, J. 1991. Les savoir-faire: une trs longue histoire.

Terrain (Carnets du Patrimoine Ethnologique) 16, 106-113.
Pelegrin, J. 1995. Rflexions mthodologiques sur ltude

de sries lithiques en contexte datelier ou de mine. In J.
Pelegrin and A. Richard (eds), Les mines de silex au No-
lithique en Europe, Documents Prhistoriques, Edition du
C.T.H.S., 7, 159-166.
Praud, I., Bostyn, F., Martial, E. and Michel, L. 2002. Un

site Villeneuve-Saint-Germain dans la valle de lOurcq.
Interno 4, 13-22.
Silvestre de Sacy, L. and Baudouin, M. 1926. La station

campignienne du Clos, Flins-sur-Seine (Seine-et-Oise).
Bulletin de la Socit des Sciences de Seine-et-Oise fasc.
4, 2e srie, t. VII, 49-62.
Silvestre de Sacy, L. and Baudouin, M. 1927. La station

campignienne du Clos, Flins-sur-Seine (Seine-et-Oise).
Bulletin de la Socit des Sciences de Seine-et-Oise fasc.
4, 2e srie, t. VIII, 50-58.
Watt, J.-P. 2007. Objets en silex tertiaire imports en

Haute-Normandie. Haute-Normandie archologique 12,
53-67.
95
96
Jerzy Bbel: The Krzemionki flint mines latest underground research 2001-2004
The Krzemionki flint mines

Latest underground research 2001-2004
Jerzy Bbel
Abstract: In 2001-2004 the new underground gallery was made in the Neolithic flint mine at Krzemionki Opatowskie,
southern Poland. The gallery passes through the area of eight shafts. The author with the team from Archaeological and
Historical Museum at Ostrowiec witokrzyski carried on the archeological invesigations during the mining works con-
nected with the preparations of the gallery. The new observations were made concerning the stratigraphy of rubble tips,
traces of works, tools, illumination and ventilation in the prehistorical shafts and so on. The new 10 radiocarbon dat-
ings from pices of charcoal were obtained. Now the uderground touristic route at Krzemionki is approxiately 460 m in
length.
Keywords: Mining field, Neolithic flint mines, flint workshops, fireplaces in mines, ventilation, charcoal drawings,
wooden torch, dry filling, wooden, stone and antler tools, wooden constructions, traces of works, traces of rainwater in
mines, underground tourist tail.
In the Mesozoic margin of the witokrzyskie Mountains (1953, 1958-1961), Jan Kowalczyk, Bogdan Balcer and
(=Holy Cross Mts.) in southern Poland, there are vast out- Zygmunt Krzak (1969-1970), Jerzy Bbel (1979-1984,
crops of various kinds of flint deposits, and many prehis- 2001-2006), Sawomir Saaciski, Marek Zalewski, Wi-
toric flint mines. Mines of flint showing the characteristic told Migal (1985-1988) and Wojciech Borkowski (1989-
striped texture have been found at Korycizna, Borownia 2000), along with many colleagues.
and Ruda Kocielna, along the Kamienna river valley,
north of the Holy Cross Mountains, NE of Ostrowiec The mining field at Krzemionki is located in an area of
witokrzyski, and also at Krzemionki, 8 km NE of Os- exposed Jurassic (Upper Oxfordian) limestone, along the
trowiec witokrzyski (Budziszewski, Michniak 1983/89). sides of a small syncline (Fig. 1). The parabola-shaped
The complex of flint mines at Krzemionki, (also known as field mine is about 4.5 km long, covering an area of circa
Krzemionki Opatowskie, in terms of area of the mining
field), is one of the largest in Europe (see: Barber et al.
1999, p.58, fig. 5.5). The perfectly preserved prehistoric
landscape and underground structure of the Krzemionki
mine are of extraordinary importance, both for archaeo-
logical investigations and for educational tourism.
The Krzemionki mine was discovered by the geologist Jan

Samsonowicz in 1922. Since then, it has been investigated
by many Polish researchers. Before the Second World War,
the excavations were directed by Zygmunt Szmit (1923,
1927), Jzef urowski (1925-1927), Stefan Krukowski
(1923, 1928-1937). After the war, excavations were con-
Fig. 1: Krzemionki Opatowskie Reserve, Kielce voivodships. Plan of
tinued by Micha Drewko (1945, 1948), Tadeusz urowski the prehistoric mining field.
97
785 000 m2. The flint concretions occur in three levels. ers and the mine from rain and snow. The miners used sets
They were exploited to a depth of up to about 9 m, with the of tools made from pieces of flint, from other rocks, and
deepest shafts dug near the hinge area of the syncline. from deer antlers. Those served as wedges, mallets, levers,
hoes and pickaxes. There was also an ingenious system to
The number of mining units is estimated at over 3,500. carry the flint output up to the surface. The miners worked
The shafts were set up 5 to 30 m apart; their depths and underground in a contracted position: half-lying, crouch-
shapes depend on the local geological conditions of the ing or kneeling. In order to save work, the excavated tun-
flint-bearing limestone level. Ball-shaped and flattened nels and chambers were only 55-110 cm high. Loosened
flint concretions were extracted in several ways, depend- limestone rubble was disposed of either on the surface,
ing mainly on the depth of their occurrence in the lime- where it was stored in characteristic heaps surrounding the
stone massif. They were excavated in shallow cavities, in shafts, or it was used to backfill the explored and aban-
niche mines (2 m deep by 4-5 m wide), in pillar-chamber doned chambers. To prevent mine roofs from collapsing,
mines (c. 4.5 m deep), and in chamber mines, up to 8-9 m pillars of solid rock were left (in pillar-chamber mines), or
deep, covering an area of c. 400 m2 or more. supports were made of limestone slabs and rubble. Fires
The mines were exploited from c. 3900 to 1600 BC (ac- made in the shafts and near their entrances provided air
cording to radiocarbon dating), by people who left arte- circulation in the mine. The mine was lit by burning resin-
facts categorized by archaeologists as belonging to the ous chips (Borkowski 1995a and 1995b, Lech 2004).
Funnel Beaker Culture (TRB), the Globular Amphorae The flint material obtained was sorted directly under-
Culture, and the Mierzanowice Culture. Mesolithic hunt- ground: only the best quality flint concretions and their
ers may have known these deposits and used the striped fragments were transported up to the surface. Outside the
flint even earlier. shaft entrance, they were sorted once more, before they
A growing Neolithic population, as well as the introduc- were subjected to preliminary shaping. The concretions
tion of the slash-and-burn type of farming, were vital were broken on a stone anvil and worked with shaping
factors that led to the development of flint mining in the tools made of stone, flint, bone and hard wood. Large
witokrzyskie region (Holy Cross). Axes made of flint, amounts of flint waste and aborted semi-products of axes
mostly used for tree-cutting, to clear land as well as to cut and other tools were abandoned near the shaft entrance
wood, were distributed over a range of 250 km around (in the site flint workshops). Selected semi-products, or
the mines at the time of the development of the Funnel roughly shaped lumps, were taken for further processing
Beaker Culture (c. 3900-2900 BC). However, most shafts to workshops located at a distance of many km from the
at Krzemionki were made by miners of the Globular Am- mine, in the basin of the Kamienna river (e.g. mielw
phorae Culture (2900-2500 BC). Gawroniec), where, for instance, axes were polished and
The axes then produced, probably for special purposes, completely finished (Balcer 1995 and 2002). Apart from
have been found as far away as 600 km from the mine. In temporary camps built by the miners, there was no perma-
the Early Bronze Age (the Mierzanowice culture, c. 2200- nent settlement in the Krzemionki mining area, because
1600 BC), tools and weapons (axes and arrow-heads) potable water was lacking. Sometimes miners used rain-
made of flint were distributed in a smaller range of c. 85 water, which gathered in some deeper karst hollows, about
km (Balcer and Kowalski 1978, Bbel J. 1980, Borkowski 250-350 m south of the mining field.
and Budziszewski 1995). After the prehistoric miners abandoned the flint deposits,
During the Neolithic Period, the development of more the area remained hidden in ancient forest until modern ag-
complex flint mining technology resulted in the growth riculture transformed the landscape, at the beginning of the
of specialization: at this time professional flint miners 20th century, when the village of Krzemionki was created
emerged. A mine crew consisted of five to ten persons. nearby. The inhabitants and the lime-producers partially
Flint was mined both in the warm season (mostly in shal- destroyed the ancient flint mines (among other things, the
low cavities) and in the cold season (in deep chamber Great Chambers in the present-day tourist route), in order
mines). to gain limestone to produce lime and to use as a flux-
Sheds were built over the chamber shafts to protect the mining agent for the Ostrowiec steelworks. This destructive
98
Fig. 2: Krzemionki Opatowskie Reserve, Kielce voivodships. So-called Trail no. 1 (acc. to J. Bbel 1990, fig. 1 ). Plan of shafts 1, 2, 3 and of loca-
tions dated with C-14. 1- solid limestone, 2- dry flling (limestone rubble), 3, 4 - modern protecting devices, 5- solid limestone removed during the
working of tourist passage, 6- tunnel route, 7- figures of prehistoric miners, 8- prehistoric drawing of a human figure (Woman giving birth to a child),
9- subterranean areas excavated in 1983 and 1984.
limestone exploitation was stopped when the archaeologi- was aided by seism-electrical and Surface Interface Radar
cal reserve was established. Its organization began in 1926 (SIR) surveys along the line crossing the main flint-bear-
(Bbel 1975). Now, the whole mining field constitutes a ing field (Borkowski 1990, Herbich 1993). At that time,
Historical Monument of Poland, and a protected Archaeo- the so-called Zenon shaft was sunk and the edge of one
logical and Natural Reserve Krzemionki Opatowskie of the chambers of mine no. 7/610 was identified. A par-
(Hadamik 2006). allel underground gallery was dug in the region of shafts
One of the important questions was how to show the pre- 1, 2 and 3 (explored in the 1950s). The opening of Trail
historic flint mines to contemporary tourists. 1 took place in June 1985 (Fig. 2, Bbel 1986). Trail 2
Thirty years ago, in 1976, I submitted to public opinion was built around mine no. 7/610 and opened to the public
the idea of displaying the Krzemionki mines by excavat- in 1990.
ing underground galleries across the site (Bbel 1983). I
had been inspired by the work carried out since 1964 in With the growth of tourism in the region around Krzemi-
the prehistoric mines at Rijckholt near Maastricht in Hol- onki, the idea of joining the trails was once again dis-
land (Engelen F.H.G., 1980; Felder et al. 1998). This pro- cussed. The Permanent Krzemionki Reserve Commission
posal was debated by specialists for several years, before (under the auspices of the General Conservation Office,
approval for its construction was granted by the Ministry Warsaw) which operated in 1999-2002, under the direc-
of Culture and Art. tion of Professor Jacek Lech, produced a programme for
The project was carried out by an excavation team from the protection and management of the reserve. One of the
the State Archaeological Museum in Warsaw, which I points in its programme was the construction of the under-
directed during the period from 1979 to 1984. Our work ground trail.
99
The fieldwork was carried out from 2001 to 2004 by the

team of archaeologists of the Historical and Archaeological
Museum in Ostrowiec, under my direction, in close coop-
eration with miners employed by the firm Geohydrowiert
from Kielce and scientists of the Metallurgy and Mining
Academy at Cracow. The establishment of the trail marked
the end of an important phase of archaeological, mining
and conservation work in the Archaeological and Natural
History Reserve Krzemionki Opatowskie.
The archaeological and mining project began with the

sinking of an exploratory shaft measuring 2-3 m x 9 m (22
m) halfway between the two tunnels, on the outer extreme
of the parabola of the mining field and beyond the limits
of the Neolithic flint workshops. In this excavation, a ven-
tilation and evacuation shaft was sunk into solid Jurassic/
Oxfordian limestone, beyond the underground workings
of the prehistoric mines (Fig. 3). Sunk in 2001, the shaft
is 11.2 m deep and reveals the complete geological sec- Fig. 3: Krzemionki Opatowskie Reserve, Kielce voivodships. Plan of
tion of the area (Fig. 4). Three flint-bearing levels were the underground tourist trail.
1- Neolithic mine unit no. 7/610, 2- Modern Zenon shaft and entry to
uncovered at the bottom of the shaft, of which the two up- the underground tourist trail, 3- Modern ventilation and evacuation shaft,
permost ones had been exploited in the Neolithic Period 4- Tunnel for tourists, 5- Great Chambers, 6- Modern Stefan shaft and
exit of the underground trail, 7- Border between solid rock and exploita-
(Fig. 5). The limestone strata incline was calculated at 3 tion field. Black points prehistoric shafts. 612, 615, 790 etc. numbers
of prehistoric shafts.
to the south-east (150 off North).
The first gallery, measuring 1.1 x 2 m, was dug in the shaft
bottom in solid limestone in the direction of the Zenon
shaft and the second underground trail (Fig. 6), taking
care to avoid the prehistoric galleries. On the way, four
windows were cut which opened only onto the edge of
the partly rubble-strewn floor of chamber mines no. 615
and 612. We did not examine any of the underground
mines exposed in this section of the trail (the ventilation
and evacuation shaft and Zenon shaft), leaving them un-
disturbed.
Large nodules of flint concretions that had been removed
during excavation in the monolithic limestone tourist gal-
lery were weighed: this showed that they are not evenly
distributed in layers, but demonstrate local density varia-
tions. The differences were as great as 30.5 kg for a length
of one metre of the gallery.
Along the course of the trail, some geological distur-
bances in the limestone strata and some vertical fractures
were observed. Dense joint fractures and fault surfaces
Fig. 4: Krzemionki Opatowskie Reserve, Kielce voivodships. Cross-
were also noticed. Joints and faults in the limestone also section of the modern ventilation and evacuation shaft. 1- humus and
subsoil, 2- sand, 3,4- clay, 5- lime rubble passing into solid rock, 6, 9- pe-
caused intense fracturing of the flint concretions. There
litic limestone, 7- coral limestone /limestone with relics of see organisms,
was no point extracting them, given the exceptionally poor 8- oolitic limestone, I,II,III flint levels.
100
quality of the material and its lack of any value for knap-
ping. That was indeed the main reason why work had been
abandoned in chamber mine no. 615 for example, which
showed small galleries and an extremely undulating roof.
It turned out to be a failed investment (Fig. 7).
In the course of our work, we reviewed seismic surveys

carried out in the 1980s. The hypothetical plan of the Neo-
lithic mines suggested at that time (Borkowski 2000: 254,
Fig. 13) does not reflect the actual state of affairs, which
seriously challenges the credibility of the seismic method
used to survey this kind of construction. In the case of deep
Fig. 5: Krzemionki Opatowskie Reserve, Kielce voivodships. Flint
Neolithic mines, seismic surveys cannot be relied upon as
nodules in situ, at the bottom of modern ventilation and evacuation shaft. the basis for archaeological interpretation. The most effec-
Photo: J. Bbel.
tive non-invasive subsurface environmental sensing tool is
Subsurface Interface Radar surveys.
After linking the gallery with Trail 2, we began the work
on a similar gallery towards Trail 1.
This section of the trail ran partly through solid limestone
and then through the rubble-strewn workings of mines no.
795, 804, 806, 815, 818 and 821 (Fig. 8).
Fig. 6: Krzemionki Opatowskie Reserve, Kielce voivodships. The mod-

ern underground tourist trail in the solid limestone rock. Photo: J. Bbel.
Fig. 7: Krzemionki Opatowskie Reserve, Kielce voivodships. Neolithic Fig. 8: Krzemionki Opatowskie Reserve, Kielce voivodships. Archaeo-
mine no. 615. Empty space near the bottom of the shaft. Photo: J. Bbel. logical underground investigation in the mine no. 795 in 2003. Photo: J.
Bbel.
101
The first mine mentioned, mine no. 795, is a chamber

mine. We chose to investigate only a fragment of its east-
ern part: two large niches and a smaller one diverge from
the main chamber, heading into the depths of the rock
mass. We initially estimated the surface area of these gal-
leries at almost 300 m. This probably amounts to less than
half of the underground chamber of mine no. 795, making
it the largest known Neolithic flint mine in Poland. Ventila-
tion is another extremely important issue. It is generally ac-
cepted that the maximum direct distance from the bottom of
the shaft to the surface should not exceed 12 m. Since a high
concentration of carbon dioxide builds up in the area where
a miner is working at the rock face, this can cause ventila- Fig. 9: Krzemionki Opatowskie Reserve. Mine no. 815. Cross-section
of the fireplace in Neolithic waste-heaps. Photo: J. Bbel.
tion problems. In the case of low passages - up to a height of
around 1 m - where torches were used, this concern was of
fundamental importance.
In chamber mine no. 795 the direct distance from the bottom
of the shaft to the furthest face was around 20 m. This face
was located at the outside boundary of the mining field, in
a place where the flint-bearing layer was dipping into the
centre of the flint syncline. In that case, suitable ventilation
was naturally present. Air warmed by the presence and activ-
ity of human bodies travelled along the ceiling towards the
shaft, while cold fresh air travelled just above the floor of the
gallery. This system worked best in winter, when there were
large temperature differences between the surface and the
underground. Please recall that the temperature underground
in Krzemionki is a constant 6-9 C throughout the year. Fig. 10: Krzemionki Opatowskie Reserve, Kielce voivodships. Mine
A great deal of charcoal (the remains of wooden torches no. 815. Neolithic charcoal drawing on the wall. Photo: J. Bbel.
which provided light during mining) was found during ex-

cavation. Its presence also indicates ventilation. In order to
create an air flow, several torches had to be burned at the
mine face and in the tunnels. An additional means of intensi-
fying the air flow was to keep a fire burning at the base of the
shaft. Vestiges of such a fire, in the form of a 10-15 cm level
of charcoal, were discovered near the spoil heap of shaft no.
815 (Fig. 9).
Torches were made of tree branches of 3-4 cm in diameter. In

order to keep the flame alight, the miner would rub it against
the sharp edge of limestone slab, a piece of waste, or against
the sidewall of the mine. This leaves behind a characteristic
pattern, a lattice of numerous criss-crossed lines. Numerous
examples of this phenomenon were observed in the cham-
Fig. 11: Krzemionki Opatowskie Reserve, Kielce voivodships. Mine
ber of mine no. 795, which was investigated. In some cases no. 804. The wooden torch in its original position. Photo: J. Bbel.
102
spoil heaps.
The gallery uncovered in mine no. 795 was made up partly
of the connecting corridor running from the bottom of the
shaft to the node and partly of the so-called technological
corridor. The purpose of the former was mainly to haul out
the excavated material. The latter led from the node to the
face uncovered in window no. 4. During the course of the
mines exploitation, the course of that gallery was modified.
The connecting corridor in mine no. 795 possesses smoothly
polished sidewalls and floor. This contrasts with the techno-
logical corridor built in the spoil heap constructed of sharp-
edged limestone nodules.
no. 795. Neolithic communication gallery. Photo: J. Bbel.
Up to now, it had been thought that the flint mined from the
rock was hauled in woven wicker baskets. With regard to
mine no. 795, this cannot have been the case, as the haulage
vessels left no clear marks or scratches on the sidewalls or
floor of the connecting gallery. A significant difficulty for
Neolithic miners would have been the tendency of such bas-
kets to get stuck in the narrow passages, or caught on the pro-
truding edges of limestone lumps, causing the gallery bottom
to be quickly covered. The baskets themselves would have
quickly been destroyed. We suggest then that the raw mate-
rial used for the production of containers was thick animal
hide. Such leather sacks, dragged on the fur side, would have
been easy to haul. The strong, flexible material would have
been only slightly deformed on bends and other obstacles,
and would have slid smoothly over uneven surfaces. This
Fig. 13: Krzemionki Opatowskie Reserve. Mine no. 795. Neolithic hypothesis should be investigated in the future by an exami-
technological gallery. Photo: J .Bbel. nation of micro traces by specialists.
The gallery in column mine no. 804 leads under the rock
the irregular lines of the marks are arranged in more regu- walls dividing up the chamber and further along the sidewall
lar patterns, which may suggest some symbolic significance to the furthest part of the mine. The sidewalls of this gallery
(Fig. 10). Future research will test this initial hypothesis.
A unique find was a well-preserved fragment of a Neolithic
torch which had escaped decay (Fig. 11). This was discov-
ered at the boundary between mines no. 815 and 804. C14
dating of part of the torch would place it in the period be-
tween 3330 and 2907 BC (Bbel et al. 2005: 542).
In the underground galleries of mines no. 795 and 804, con-
necting and so-called communication and technological
tunnels were discovered (Fig.12, 13). The gallery in one
place revealed a so-called node, while in another - mine
no. 804 - it cut straight across a Neolithic gallery. In no case
does the most recently dug tourist trail run along the pre-
no. 795. Cross-section of the prehistoric underground heap of limestone
historic gallery, but through solid rock or through Neolithic rubble used as filling material. Photo: J. Bbel.
103
Fig. 15: Krzemionki Opatowskie Reserve. Mine no. 795. Cross-section

of prehistoric underground waste-heaps (dry filling). Photo: J. Bbel.
Fig. 18: Krzemionki Opatowskie Reserve. Mine no. 806. Negative of

a wooden construction in Neolithic waste-heaps (dry filling). Photo: J.
Bbel.
are not polished and its floor was strewn with large fragments
of limestone waste. Most probably this corridor served as a
Fig. 16: Krzemionki Opatowskie Reserve, Kielce voivodships. Mine connecting artery, linking shaft 804 with the other galleries
no 795. Negative of a wooden tool in Neolithic waste-heaps (dry filling).
adjacent to this mine from the east.
Photo: J .Bbel.
The tourist tunnel that we dug out cut across several un-
derground Neolithic spoil heaps (Fig. 14, 15), which, after
extensive documentation, were suitably protected from de-
struction and collapse. The perfectly visible artificial bed-
ding produced in the limestone waste during its collection
makes possible the full reconstruction of the sequence of
work of Neolithic miners in that part of the mining area.
This issue will be the subject of a separate paper.
Fig. 17: Krzemionki Opatowskie Reserve. Mine no. 795. Positive of a During the exploration of the underground spoil heaps, we
wooden tool (in gypsum). Photo: J. Bbel.
came across small hollow spaces which were impressions
made by wooden tools and objects (Fig.16). Rotten re-
mains strewn about attest to this. Rock dust and limestone
waste sticking to them preserved the shapes of these ob-
jects. The cylindrical hollow spaces were around 50-70 cm
in length with a diameter of 3-4 cm. They were probably
the remains of levers used to prise off large slabs from the
roof of the mine. A plaster cast was made of one of these
Fig. 19: Krzemionki Opatowskie Reserve Kielce voivodships. Mine no
finds. They must also have been used as wedges to split the
795. Fragments of the mines stone pick. Photo: J. Bbel.
rock (Fig. 17). The impression of one such tool was 19 cm
104
long around 6 cm wide and up to 2 cm thick.

The impressions were sometimes found together in groups.
In two places some distance apart from each other, in mine
no. 804, we came across several impressions of wooden
poles arranged vertically and horizontally. They were 1 m
long and 2-9 cm in diameter. In one place they were clearly
located at the border of two layers of limestone waste, be-
tween a spoil heap and a workface later buried in waste.
This suggests that various wooden constructions were built
to protect the gallery from the spoil heaps collapsing, with
the result that the gallery became narrower and moving
through it became more difficult. In places where the par-
Fig. 20: Krzemionki Opatowskie Reserve Kielce voivodships. Mine no.
ticular type of rock mass split into small cubes, it was im- 795. Neolithic antler tool in its original position (in situ). Photo: J. Bbel.
possible to produce large limestone slabs to line the corri-
dor, and primitive timbering was put up instead (Fig. 18).
Plant remains, which may have derived from woven

mats, have been found among the waste in underground
Neolithic spoil heaps. Plaited mats may have also been Fig. 21: Krzemionki Opatowskie Reserve Kielce voivodships. The ant-
ler wedge. Photo: J. Bbel.
used during mining to insulate the body of a crouching
or prone miner from the cold and damp rock floor. This
would have been a natural measure to prevent such ill-
nesses as rheumatism, gout or acute lumbago etc.
Other mines, encountered during the connection of the
tourist trails, much smaller and shallower than mine no.
795, can be categorised as chamber and pillar mines.
The examination of the underground spoil heaps also re-
vealed well-preserved fragments of stone and flint tools
discarded by miners (Fig. 19). The majority of them are
cigar-shaped picks made of crystalline rock, and ham-
merstones. The latter were mainly used as hammers to
test the quality of the raw material extracted from the Fig. 22: Krzemionki Opatowskie Reserve, Kielce voivodships. Mine
no. 805. Traces of works on the wall of Neolithic gallery. Photo: J. Bbel.
rock. We also found wedges made of fragments of pol-
ished axes, characteristic of the Globular Amphora cul-
ture. Another group of tools made of fragments of red
deer, roe deer and elk antlers were used as wedges, le-
vers, picks and gouges (Fig. 20, 21). A spade made from
an animals shoulder blade, found in one of the work
faces of mine no. 795, was probably used to rake up and
scatter fine waste.
On the mine sidewalls, and particularly on work faces

in niches, thousands of marks left by the tools of Neo-
lithic miners can be seen (Fig. 22). These marks provide
us with a great deal of information about the miners
themselves. They were right-handed. They would change no 815. Traces of work of an antler gouge. Photo: J. Bbel.
105
tools, depending on the work being carried out at the time

and on the type of rock they had to deal with in a given
location, in a remarkably rational way. The best exam-
ples are the marks of horn wedges used to prise off large
pieces of limestone on the roof of mine no. 815 (Fig. 23).
Gouges used there left indentations of their blades in the
clay which filled a karst crevice. The technique of ex-
tracting flint nodules from rock is illustrated by a series
of marks in one of the faces in mine no. 795 (Fig. 24).
There, marks of impacts can be seen which split off the
concretions from the rock sidewall, as well as the mark of
a lever which split flint away from the floor. In mine no.
795, marks of where large limestone slabs were broken Fig. 24: Krzemionki Opatowskie Reserve. Mine no 795. Negative of a
flint nodule. Photo: J.Bbel.
off from the roof of the workings can be seen (Fig. 25).
The size of these slabs and their weight - which reached as
much as several hundred kilograms in some cases - force
us to consider how they were broken off in such cramped
and low passages.
Large accumulations of flint flakes together in a niche of
pillar mine no. 818 argue for the existence of an under-
ground workshop for preliminary flint knapping, to test the
stones quality. Other places like these were discovered in
the region of Trail 1 in 1983.
Before the flint reached the workshop on the surface next

to the shaft, its use value was assessed. Micro cracks,
formed as a result of tectonic movements of the rock mass,
make up part of any normal concretion. Neolithic miners,
working underground by the light of a torch, would obvi-
Fig. 25: Krzemionki Opatowskie Reserve. The mine no. 795. Traces of
ously not have been able to see such cracks; nonetheless,
works of a wooden wedge. Photo: J.Bbel.
most poor quality flint remained below the surface. In my
opinion, the sense of hearing rather than of sight played a
crucial role in testing flint. As he worked, the miner not
only saw the pieces of flint breaking and falling under
blows of his hammerstone, but he also heard whether the
sound was clean or dull. In the latter case, the flint was of
no further use and ended up along with limestone waste on
the underground spoil heap.
The horizontal watermark of a temporary underground

lake can be seen on the walls of Neolithic galleries built
from limestone slabs, and on the sidewalls of the deep-
est part of mine no. 795. It must have arisen as a result of
heavy rains before the shaft was buried under waste. Rain-
water entered the underground from the surface through
no 795. Traces of water filling on the limestone rubble in the chamber-
the open shaft (during the Neolithic Period?) carrying in mine. Photo: J. Bbel.
106
The tourist trail has been made completely safe for visitors
(Fig. 27). Traditional support techniques have been used,
along with pressure injection methods. In addition, Neo-
lithic man-made spoil heaps have been fenced off from the
public by steel mesh hung between columns to prevent un-
desirable interference.
The underground exhibition gallery, about 480 m long,
passing through Neolithic mining units, was opened for
tourists on l July 2004.
The Krzemionki underground tourist trail has great edu-

cational value. In 2004-2006, over 160 000 people visited
Fig. 27: Krzemionki Opatowskie Reserve, Kielce voivodships. Mine it, making it a record in the history of the museum and
no. 795. The underground tourist trail is ready. Photo: J. Bbel.
archaeological reserve.
In May 2005, the Minister of Culture awarded the Historical
with it various soil components, mud, flecks of humus and and Archaeological Museum in Ostrowiec witokrzyski
of rotten vegetation. Before the water soaked into the rock, the third prize in the category of archaeological exhibitions
the scum on its surface left its mark in the form of a thick in the competition of Museum Events - Sybilla 2004, for
line (Fig. 26). Identical marks can be seen on the Neolithic its underground tourist trail created in the Neolithic flint
sidewalls of the galleries in the Great Chambers. These mines of Krzemionki.
observations may interest climatologists researching cli- And in October 2006, the Minister of Economy awarded
mate changes in past epochs. our Museum the Certificate from the Polish Tourist Organ-
isation Best tourist product of 2006 Travel with passion
During the excavation, numerous samples of charcoal year for the underground tourist route in prehistoric flint
were extracted from the part of the trail running through mines and the reconstruction of the Neolithic village in
chamber mines and chamber-and-pillar mines. Thanks to the Krzemionki reserve.
a collaborative agreement with the Holy Cross Academy Another extremely important benefit is that the tunnel has
of Kielce and the University of Erlangen in Germany, we opened space for future specialised research. This, with the
obtained 10 new C 14 datings (Bbel et al. 2005). After help of improved techniques and technology, will signifi-
calibration, these gave readings of around 3500 - 2900 cantly improve the current understanding of the prehistory
BC. Previously, it had been thought that chamber mines of the Krzemionki mines.
were the work of miners of the Globular Amphora culture.
These dates however are inconsistent with the generally
accepted chronology for that culture. This means that the Dr Jerzy T. Bbel
question of during which culture these mines were exploit- ul. M. Dabrowskiej 19m9
ed remains controversial. 01-903 Warszawa POLAND.
e-mail: tbabel@poczta.onet.pl
A long stretch of the tourist trail runs beyond the mining website: www.krzemionki.pl
area along a gallery dug out of solid limestone on whose
sidewalls the section of flint-bearing layers can be seen.
Observations of these sections and planigraphy aided ge-
ologists from the Polish Geological Institute to propose
a new hypothesis for the genesis of Jurassic (Oxfordian)
banded flint (Piekowski and Gutowski 2004).
107
References striped flint in prehistory). Wiadomoci Archeologiczne,

43(2), 127-145.
Bbel, J.T. 1975. Zniszczenia, badania i ochrona rezer-
watu w Krzemionkach. Wiadomoci Archeologiczne 40, Barber, M., Field, D., and Topping, P., 1999. The Neolithic
149-177. Flint Mines of England. London.
Bbel, J. 1980. Krzemionki. In G. Weisgerber, R. Slotta, Borkowski, W. 1990. Results of subsurface radar geophysi-
and J. Weiner (eds), 5000 Jahre Feuersteinbergbau. Die cal studies of the Krzemionki banded flint mines Poland.
Suche nach dem Stahl der Steinzei, Bochum, 86-595.. Archeometry 90, 687-696.
Bbel, J. 1983. Krzemionki dzi i jutro. Stan, badania i Borkowski, W. 1995a. Krzemionki Mining Complex:
perspektywy zagospodarowania rezerwatu archeologic- deposit Management System. Warsaw
znego. Wiadomoci Archeologiczne 48, 223-236.
Borkowski, W. 1995b. Prehistoric flint mines complex in
Bbel, J. 1986. The Problems of investigations of the flint Krzemionki (Kielce Province). In Z. Kobyliski and J.
mine at Krzemionki, near Ostrowiec witokrzyski. In Lech (eds), Archaeologia Polona 33, 506-24.
K.T. Biro (ed.) International Conference on Prehistoric
Flint Mining and Lithic Raw Material Identification in the Borkowski, W. and Budziszewski, J. 1995. The use of
Carpathian Basin, Budapest - Smeg, 20-22 May 1986, stroped flint in Prehistory. In Z. Kobyliski, and J. Lech
Budapest, 27-42. (eds), Archaeologia Polona 33, 71-87.
Bbel, J. 1990. The flint mine at Krzemionki and the prob- Borkowski, W. 2000. Zastosowanie metody radarowej
lem of flint workshop from the Early Bronze Age in the i sejsmicznej w lokalizacji podziemnych wyrobisk gr-
Central Little Poland. In M.-R. Seronie - Viven and M. niczych na podstawie wynikw prac prowadzonych w
Lenoir (eds), Cahiers du Quaternaire no 17 - Le silex de sa Krzemionkach Opatowskich. In W. Borkowski (ed.), Me-
gense loutil, Actes du Vme colloque international sur le tody bada archeologicznych stanowisk produkcyjnych
silex. Bordeaux, 201-209. grnictwo krzemienia. Warszawa.
Bbel, J., Braziewicz, J., Jaska, M., Kretschmer, W., Pa- Budziszewski, J. and Michniak, R. 1983(89). Z bada nad
jek, M., Semaniak, J., Scharf, A. and Uhl, T. 2005. The wystpowaniem, petrograficzn natur oraz prahistoryczn
radiocarbon dating of the Neolithic flint mines at Krzemi- eksploatacj krzemieni pasiastych w poudniowym skrzy-
onki in Central Poland. Nuclear Instruments and Methods dle niecki Mago Folwarczysko. Wiadomoci Archeo-
in Physics Research 240, 539-543. logiczne 49, 151-189.
Balcer, B. 1995. The relationship between a settlement and Engelen, F.H.G. 1980. Rijckholt-St.Geertruid, Prov. Lim-
flint mines. A preliminary study of the Eneolithic work- burg. In G. Weisgerber, R. Slotta and J. Weiner (eds), 5000
shop assemblages from mielw (Southern Poland). In Z. Jahre feuersteinbergbau. Die Suche nach dem Stahl der
Kobyliski, Lech J. (eds), Archaeologia Polona 33, 209- Steinzeit, Bochum 559-567.
221.
Felder, P.J. (Sjeuf), Rademakers, P., Cor, M. and de Grooth,
Balcer, B. 2002. mielw Krzemionki- wieciechw. M. E. Th. (eds.) 1998. Excavations of Prehistoric Flint
Zwizki osady neolitycznej z kopalniami krzemienia. Mines at Rijckholt - St. Geertruid (Limburg, The Neth-
Warszawa. erlands) by the Prehistoric Flint Mines Working Group
of Dutch Geological Society, Limburg Section. Archaeolo-
Balcer, B., and Kowalski, K. 1978. Z badan nad krze- gische Berichte, 12, Bonn.
mieniem pasiastym w pradziejach (Sum.: Studies on
108
Hadamik, Cz. 2006. Krzemionki- Flint Mines from the

Neolithic Period. In M. Rziewicz and J. Wendland (eds),
Historic Monuments in Poland, 30 Treasures of National
Heritage, Warszawa, 140-146.
Herbich, T. 1993. The method of estimation of the extent

of the mining field of flint mines through observation of
the arrangement of surface layers. Archeologia Polski 38,
23-35.
Lech, J. 2004. O badaniach prehistorycznego grnictwa

krzemienia i kopalni w Krzemionkach Opatowskich.
Przegld Archeologiczny 52, 15-88.
Piekowski, G. and Gutowski, J. 2004. Geneza krzemieni

grnego oksfordu w Krzemionkach Opatowskich, Genesis
of the Upper Oxfordian flint in Krzemionki Opatowskie,
Poland. Tomy Jurajskie 2, 29-36.
109
110
Laurence Manolakakis: Open-cast flint mining, long blade production and long distance exchange: an example from Bulgaria
Open-cast flint mining, long blade production

and long distance exchange:
an example from Bulgaria
Laurence Manolakakis
Abstract: In the early 5th millenium, a vaste area termed Graphite Pottery Zone in the Balkan Peninsula underwent
radical transformation: emergence of hereditary social hierarchy, specialization and regionalization of productions, highly
organized long distance exchange, and technological innovations. A particular aspect of the latter was the specialized pro-
duction of very long flint blades. In north-east Bulgaria, Ravno near Razgrad, is a good-quality flint deposit, and the tell
of Kamenovo is a characteristic village of the Graphite Pottery Zone, with two major differences: the village is located on
a flint deposit and the inhabitants produced long blades by lever pressure as prestige items. The raw material procurement
of Kamenovo combines exploitation of flint on the site for domestic production and further exploitation a short distance
away (Ravno) of high quality raw material for specialized production. Settlement location away from the Ravno source
may reflect the fact that Ravno was probably exclusively exploited by specialists, maybe as a protected area, and that long
blade debitage was not an intensive production. One may suggest that intensive exploitation developed when an object
was not or no longer a prestige item.
Keywords: Open-cast flint mining, domestic production, specialization, exploitation, procurement, Eneolithic, Bulgaria.
1. Cultural-historical context
From the early fifth millennium BC, certain populations minology (Lichardus et al. 1985). While the end of this
in Europe underwent radical technological, economic and era is poorly dated and documented, there is abundant
social changes. The term Chalcolithic is usually applied evidence throughout the first half of the fifth millennium,
to these upheavals, which included the establishment of allowing definition of a multitude of cultural groups, cul-
durable social hierarchy, the emergence of virtually pro- tures, regional groups and cultural complexes (Todorova
fessional craftsmen, and the development of long-distance 1978). Yet the various cultural groups which made up the
networks for circulation and exchange of goods. The cra- Graphite Pottery Zone share a number of features, particu-
dle of this transformation was the Balkan Peninsula, with larly flint production.
a major cultural area called the Graphite Pottery Zone
covering the whole of Bulgaria, the Romanian Danube A clearly differentiated, hereditary social hierarchy, which
plain, the Republic of Macedonia and the north coast of first appeared in the early Eneolithic, was completely es-
the Aegean. It lasted from about 5000 to 4300 BC and is tablished in the late Eneolithic, as reflected in the Varna
divided into two main periods. According to the terminol- cemetery (Ivanov 1978). In the early Eneolithic, copper
ogy used in Bulgaria, based on the tell of Karanovo, these metallurgy not only involved the manufacture of small
are the early and late Eneolithic, corresponding to the late objects but also of heavier tools such as axes and ham-
Neolithic and early Chalcolithic of the pan-European ter- mer-axes. In the late Eneolithic gold-work emerged, for
111
3 4
2 Fig. 1: Some samples of blades from the domestic and specialized

debitages in the North-East, Early and Late Eneolithic.
5 6 7 8 9 10 11 12
112
both ornaments and symbols of power, particularly scep- of the five regions, the same debitage was used, indicat-
tres. The whole Eneolithic is also characterized by high- ing that all settlements shared the same technical knowl-
temperature firing of pottery decorated with graphite, or edge and conception of debitage. In the current state of
with gold in the cemeteries, by lever pressure debitage, research, there is no evidence for specialized production
and long-distance circulation networks for certain goods within the domestic sphere. Regional domestic production
such as copper and very long flint blades. can be considered traditional, as it was already present
in the early Eneolithic and remained unchanged in the
Production was organized regionally, in accordance late Eneolithic. A roughly east-west separation is visible,
not only with the presence of raw materials copper, between the north-east and Thrace on the one hand and
graphite, spondylus shell, gold, flint but also with tech- west Bulgaria on the other. Briefly put, indirect percus-
nical know-how, as shown by the various production cen- sion is used in the eastern area and direct percussion in
tres for very long flint blades (Manolakakis 2005), gold the west. Due to its considerable uniformity, the eastern
ornaments (Angelov 1959, Hartmann 1978), copper tools half of Bulgaria (north-east and Thrace) formed a large
(Tchernih 1978a & b), bone figurines (Angelov 1958, unit matching the cultural complex defined on the basis of
1961) and probably graphite decorated pottery as well ceramics, the Boian-Marica of the early Eneolithic and the
(Demoule 2004). Kodzadermen-Gumelnitsa-Karanovo VI of the late Eneo-
lithic, both belonging to the Graphite Pottery Zone.
2. Lithic productions
At the same time, there is specialized debitage: throughout
Lithic production using flint can be summarized as fol- the periods, there is evidence from all the villages in the
lows (Manolakakis 2005) (Fig. 1 n5-8): domestic blade eastern half and some in the western half of a few blades
debitage was carried out within each village, using local which do not originate from domestic debitage, and for
flint. This occured in all regions and represented a large which there is no knapping waste present. These blades al-
majority of cases (90% of lithics on a site). Products ways represent less than 12% of the lithic industry from
did not circulate between villages, or between regions. villages, but they are much more common as grave-goods
Debitage differed from one region to another. Within each in cemeteries, where they represent as much as 40 to 68%.
This is mainly lever pressure debitage of very long
blades, from 22 to 45 cm long. To a lesser extent,
crutch pressure debitage occurs, of shorter blades,
roughly between 16 and 20 cm long (Fig. 1 n1-4
and 9-12). These objects circulated between vil-
lages as semi-finished products, up to a distance of
700 km .
During the late Eneolithic, though only in the east-
ern zone, axes made from very long blade cores
also circulated, as finished products and in smaller
quantity (Fig. 2).
3. Deposits and raw material ex-

traction
The raw material used for this pressure debit-

age is exclusively Aptian flint from the so-called
1 2
pre-Balkanic platform in north-east Bulgaria
(Natchev et al. 1981, Natchev, Kantchev 1984)
Fig. 2: Flint axes made on pressure lever cores, only in the Late Eneolithic. (Fig. 3). There, this flint was also used for domestic
113
Fig. 3: Map of the siliceous rocks of Bulgaria (from Natchev et al. 1981) and geological location of the study area.
debitage. While all the north-eastern villages carried out Thus in the small Topciiska valley, a dozen geological
indirect percussion debitage on local flint, only a few trenches up to 5 m deep were opened, and a surface-area
used pressure debitage as well. However, there was un- of 543 ha surveyed.
til recently no evidence either for deposits containing flint
of an appropriate size for debitage of long blades, nor for Three deposits were located at Ravno, containing abundant
production sites. Debitage of very long blades, of course, raw material. The interest of this area was further enhanced
implies nodules not only of good quality but also of large by the presence of a late Eneolithic settlement at Kamen-
dimensions. ovo (Fig. 4).
Lower Cretaceous flint is abundant throughout the Moe- Ravno 3 is the only deposit which provides nodules suit-
sian plateau, where Hauterivian and Barremian flint con- able for debitage of long blades; the site is part of a regional
taining deposits are the most frequent. But there are also system of dolines and lies on a slope which flattens out
small concentrations of weathered Aptian flint, especially towards the bottom. Four trenches were lined up from the
north of Razgrad (Fig. 3) (J. Delpine, in unpublished ex- summit down to the bottom, enabling complete coverage of
cavations report 1996). geology and sediments. They were 3-15 m long, 1 m wide
with a depth ranging from 3-5 m. The stratigraphic section
This is why geological and archaeological survey of the of the trench at the bottom of the slope (Fig. 5)shows:
region between Kubrat and Razgrad was undertaken from - At the top, superficial silts.
1994 to 1997 in collaboration with Ivan Ivanov (Muse- - At the base, the Lower Cretaceous limestone bedrock,
um of Varna) and Jacques Delpine (geologist). The two dating to the Hauterivian and the Barremian, which can be
main objectives of the project were to locate flint deposits up to 500 m thick.
with nodules large enough for lever pressure debitage and - Between the two are the bright red or yellow clays of the
to find archaeological sites with waste from this kind of residual, weathered Aptian layer, containing large num-
debitage. bers of flint nodules.
114
Fig. 4: Topographic location of Ravno deposit and the tell of Kamenovo (Delpine, unpublished excavations report 1996).
Fig. 5: Stratigraphical profile of the lowest geological trench of Ravno 3 deposit (Delpine, unpublished excavations report 1996).
115
The test was made with direct percussion with a hard ham-
mer (probably in stone). In most cases, the shape of the
nodule meant that the first flake also created the future
striking platform. In other cases, the nodule was tested
by removing flakes on two opposite sides, using a natu-
ral striking platform. Raw material of poor quality was
immediately identified by most tests. However, some ex-
amples indicate that an attempted preparation had been
abandoned due to cleavage within the nodule. Most cores
were pyramidal with flat striking platforms and a wide
debitage surface, 15-18 cm long. Debitage was of long,
wide blades, curved and varying in regularity. The cores
were shaped by means of crests, often leaving the back
more or less cortical. There were also cone-shaped cores,
with very oblique and flat striking platforms, narrow
debitage surfaces and pointed bases. They are 10-15 cm
long and indicate debitage of medium-sized, slightly
curved regular blades. There were considerable numbers
of large first flakes, shaping flakes, crested blades, core
rejuvenation tablets, debitage splinters, debitage flakes,
laminar flakes for renewing debitage surfaces, and crested
flakes. Most blades had flat butts, abrased overhangs, dif-
Fig. 6: Stratigraphical profile of the shaft of Ravno 3 flint deposit fuse bulbs, two or three regular or irregular arrises, and
(Delpine, unpublished excavations report 1996).
generally a curved profile. These blades are quite short (9
cm). The longer ones (14 cm) are in fact laminar flakes
The nodules, of good quality, are 40 to 80 cm long and from debitage surface renewal (correcting hinged fracture).
occasionally reach 1 m. They are globular in shape, some- The blades are often fragmentary. The finds from the shaft
times with small surface lumps. The fresh flint is brown to thus present all stages of debitage, probably carried out by
beige in colour and the cortex is non-effervescent white. indirect percussion, and logically consist almost entirely
of waste from these stages. The shaft was indeed used to
An archaeological feature was discovered in the trench at extract flint. The preliminary stages of shaping were car-
the bottom of the slope. The shaft, which appeared just ried out by direct percussion with a hard hammer, then full
below the topsoil, is funnel-shaped with a wide opening debitage by indirect percussion. This is no evidence to date
at the top (Fig. 6). At a depth of about 2.5 m, the feature the site to the late Eneolithic, as punch percussion debitage
rapidly becomes a round shaft with an average diameter also took place during the following period (Bronze Age
of 0.60 m. The depth is over 4.5 m, since the base was according to Bulgarian terminology), and no pottery was
not reached. So narrow a feature suggests that this was an found. One can also note that the cores and waste from
access to a gallery in the flint bed. The bed, located at a the shaft are much less regular than on Eneolithic sites in
depth of 2.50-5 m, depending on the position on the slope, the north-east. Additionally, in the late Eneolithic, indirect
contained large numbers of quite large flint nodules. The percussion debitage was always carried out in the settle-
fill of the shaft consisted of speckled clayey silt containing ments themselves. Since there are no traces of settlement
a great quantity of worked flint, although there is no waste at Ravno, this would suggest that the material is posterior.
from lever pressure debitage. In this situation, it was important to investigate the small
The untouched nodules are all smaller than 12 cm. Large late Eneolithic tell of Kamenovo, about 2 km away, espe-
numbers of nodules had been tested but rejected because cially as it lies on a large flint deposit.
of poor quality: frost damage, cleavage or other defects.
116
4. The tell of Kamenovo - Level 0: 0.80-1.10 m thick, light grey, clayey silt with
some grey-brown lenses of similar texture. This level was
The tell is located next to the Topciiska, at an altitude of disturbed and the archaeological finds here are out of con-
226 m in the same geological context as Ravno (Fig. 4). text. Micro-sediment analysis should determine whether
Large quantities of good-quality flint are available on the this was a backfill horizon. The presence of this disturbed
site, but the nodules are smaller than at Ravno 3. In 1999, level is perhaps due to the location of the trenches quite
in collaboration with Ivan Ivanov and Aline Averbouh, low down on the side of the tell. Although there were
two perpendicular trenches were opened, 20 m long and clearly no features in this disturbed level, finds were very
2 m wide (Fig. 7, trenches A, B). These trenches could abundant and are all datable to the Eneolithic.
not be placed on the top of the tell as this is occupied by - Level 1: appeared at 1.60-1.80 m depth, fine clayey grey.
the present-day village. They were thus located in the only This first undisturbed level was only found in the north-
accessible area (a public garden), on a lower-lying part of east/south-west trench (about 18 square m, trench A) and
the tell. The project was interrupted after the death of Ivan produced two archaeological features. Feature 1 was vis-
Ivanov, but provided interesting results. Most interpreta- ible for only 0.60 m within the excavation trench. It was a
tion of the data remains hypothetical because of the limited roughly rectangular mass of burned daub, apparently part
scale of the excavations. The basic stratigraphy can be de- of the wall of a collapsed building. Similarly, feature 2 was
scribed as follows (Fig. 7): visible in the section and for 0.80 m within the trench. This
- Topsoil: 0.20-0.40 m thick, clayey brown, containing too was a more or less rectangular mass of burned daub
some Eneolithic ceramics and lithics, as well as a few ob- and probably also a fragment of collapsed wall. As both
jects from the Thracian period and a great many from the features were mainly seen in the profile, it was not pos-
19th and 20th centuries. sible to determine whether they corresponded to a single
structure or to two adjacent buildings.
Fig. 7: Topography of the tell of Kamenovo, location of the archaeological trenches A and B (upper part), eastern stratigraphic profile of the trench A
(lower part).
117
4 5
6 7 8
9 10
Fig. 8: Some samples of the cores from the three different knapping techniques at Kamenovo (Level 1).
118
Finds from level 1 are abundant and uniform. The shapes Lastly, all the elements of the production sequence for
and decoration of the pottery are characteristic of Kara- crutch pressure are present (Fig. 8 n2, 3), showing that
novo VI, late Eneolithic: lids with handles, carinated ves- this too was carried out on the site. The crutch pressure
sels, flat bases, decoration with graphite, barbotine and cores are short, cylindrical with peripheral debitage, or
incisions etc. (Ivan Ivanov, in unpublished excavations flat with a single debitage surface. The blades have the
report 2000). A ladle and a feeding-bottle are also part of same marks as the lever pressure ones but are narrower
the inventory. As well as pendants, pins, perforated discs, and shorter (Fig. 8 n5, 7, 8). This is the least abundant
sawn and abraded astragali, the numerous bone and antler debitage. In fact, large nodules are not required and this se-
tools include antler punches that were probably used for quence by crutch pressure could in theory have alternated
indirect percussion debitage (Aline Averbouh, in unpub- with lever pressure on the same cores. But the shape of the
lished excavations report 2000). There are very few hard- cores with crutch pressure debitage is not really compara-
stone finds, including a few small flakes and two small ble to the cores with lever pressure debitage. Therefore, it
axes in undetermined green or black rock. is difficult to determine whether flint for crutch pressure
was acquired on or outside the site.
There are large numbers of flint artefacts, especially in the
disturbed level (over 20,000 objects). In level 1, however, The flint tools are of the kinds usually found on settle-
there is much less debitage waste and a greater quantity of ments of this period (Fig. 9): scrapers on blades (n 13,
full debitage blades and tools. This would appear to sup- 14) and a few on flakes (n 15), borers and drills (n4, 5,
port the hypothesis that level 0 was a backfill. 6, 9) unmodified used blades, sickle blades (n 1-3, 10-
Amongst the thousand or so lithic artefacts from level 1, 12), retouched blades (n 8) and some retouched flakes.
punch percussion debitage is the most common (Fig. 8 n1, It should be noted that there are a considerable number of
4). The cores are globular, flat or pyramidal, with a flat tranchet axes (Fig. 9 n 17,18), unlike most settlements in
unique striking platform and posterior-lateral crests. The the north-east; elsewhere, these differences in the relative
back can remain completely cortical or can be modified. frequency of tools have been interpreted in terms of pref-
The profiles of blades are curved or very curved. They are erential activities on each site.
wide, with more or less regular arrises. Many blades have
a flat butt, with a diffuse bulb and the overhang removed. 5. Hypotheses on the modalities of
The whole production sequence is present: from tested Eneolithic exploitation and procure-
ment
and discarded nodules to full debitage blades, as well as
blade and flake tools. Given the presence on the site itself
of nodules of adequate size, this is clear evidence for flint Thus we are clearly in a settlement context where, as
acquisition and debitage at the settlement. throughout the Graphite Pottery Zone, the inhabitants of
Kamenovo carried out their domestic production by punch
Specialized debitage using lever pressure is also attested percussion. But there are two major differences:
(Fig. 8 n6, 9, 10), but less frequent. The lever pressure - They located their village on a flint deposit.
cores are long, subrectangular in section, with a flat pres- - They also carried out specialized production using lever
sure platform and straight profile. The back is modified and crutch pressure.
and the sides are prepared with crests. The blades, all frag- Yet this specialized production did not exploit the deposit
mentary, have a very small flat butt, an outstanding bulb, on which the village was located.
a carefully removed and abraded overhang, and regular, As for the forms of raw material exploitation, the inhab-
straight ridges (Fig. 8 n10). There are large trimming and itants of Kamenovo chose to position their village on a
shaping-out flakes, crested blades, abandoned cores, a few source of good quality raw material, but they also chose
fragments of large blades, tools, and axes made on cores. the spot where the flint layer is directly accessible at the
These finds, together with the absence of very large nod- foot of the slopes, in the gully cut by the Topciiska.
ules, indicate that raw material was acquired outside the Deep mine-shafts and galleries were dug elsewhere in the
site, although debitage took place on the site. Graphite Pottery Zone, to extract copper ore.
119
2
1 3
4 5 6 7 8 9
10 11 12
13 14 15
17 18
16
Fig. 9: Tools from Kamenovo (Level 1) : Sickles, borers, scrappers, flint axe, tranchet axes.
If one accepts the later date for the shaft at Ravno 3, one an intensive production. This may have been another rea-
can see here, in the case of flint, that these Eneolithic pop- son to establish the settlement away from the source.
ulations took good advantage of what was available from a
simpler form of access. The very long blades reflect highly skilled production tech-
niques: the blades were prestige goods of great social and
Concerning raw material procurement, there were prob- cultural value in the Graphite Pottery Zone (Manolakakis
ably two distinct modes, depending on the kind of produc- 2006). Three points should be recalled here:
tion: - These blades occurred frequently in funerary contexts,
- Firstly, exploitation on the site, at Kamenovo, of flint where they represented 40 to 70% of lithics.
used mainly for domestic production; - The longest blades were associated with the richest
- Secondly, exploitation at a short distance, probably at graves.
Ravno 3, of raw material that was of exceptional quality, - The blades circulated over long distances and were
at least for its nodule size, and intended for the specialized present throughout the Graphite Pottery Zone, acting as a
production of very long blades. form of barrier against the obsidian which was widespread
As for debitage using a crutch, the procurement mode re- in all the neighbouring cultures.
mains unknown: either from Kamenovo or from Ravno 3.
- Thus the two particularities of Kamenovo, specialized Clearly, there is no structural link between a highly skilled,
debitage and location on a flint source, do not seem to be socially valued production and a mode of exploitation in-
connected in a cause and effect manner. volving intensive production, deep mine-shafts and galler-
- The people did not settle there to make very long blades. ies.
They chose to settle at a short distance from the biggest One could even reverse the question: is it rather when an
flint nodules, and this probably means that the Ravno 3 object is not or is no longer a prestige item that intensive
source was exclusively exploited by the specialists, as a or industrial exploitation develops?
protected area.
Everything suggests that the specialized debitage was not
120
Laurence Manolakakis Manolakakis, L. and Ivanov, I.S. (eds.) 1995-1999. Les

CNRS-Protohistoire Europenne-UMR 7041 ateliers de taille du silex de Bulgarie du Nord-Est. Unpu-
MAE, 21 alle de lUniversit blished excavation Reports. Mission franco-bulgare, Mi-
92023 Nanterre, France nistre des Affaires Etrangres.
e-mail: laurence.manolakakis@mae.u-paris10.fr
Manolakakis, L. and Ivanov, I.S. (eds.) 2000. Les fouilles
du tell de Kamenovo, habitat et atelier de taille chalcoli-
References thique. Unpublished excavation Report. Mission franco-
bulgare, Ministre des Affaires Etrangres.
Angelov, N. 1958. Selishtna mogila pri s. Hotnica. Studia
in Honorem Acad. D. Detchev 390-403. Natchev, I., Kovnurko, G. and Kantchev, K. 1981. Kremac-
nite skali v Balgarija i tjahnata esploatacija, Interdiscipli-
Angelov, N. 1959. Zlatnoto sakrovishte ot Hotnica. Arhe- narii Izsledvanija. Sofia: Arheologiceski Institut i Muze
ologija Sofia 1, 38-46. na BAN, VII-VIII, 41-58.
Angelov, N. 1961. Rabotilnica za ploski kosteni idoli v Natchev, I. and Kantchev, K. 1984. Aptian and quaternary
selishtnata mogila pri s. Hotnica, Tarnovsko. Arheologija flint in Noerth-East Balgaria, Report of the III seminar in
3, 34-38. Petroarchaeology, Plovdiv, 65-82.
Demoule, J.P. 2004. La cramique du Nolithique Rcent. Tchernih, E.N. 1978. Aibunar, a Balkan copper mine of
In R. Treuil (ed.), Dikili Tash, village prhistorique de the fourth millenium BC. Proceedings of the Prehistoric
Macdoine orientale (Fouilles Jean Deshaye 1961-1975), Society 44, 203-217.
Ecole Franaise dAthnes. Paris: De Boccard, 63-270.
Tchernih, E.N. 1978. O jugo-vostotchnoi zone Balkano-
Hartmann, A. 1978. Ergebnisse der spektralanalytischen Karpatsko metallurgitchesko provincii epohi eneolita. In
Untersuchung neolithischer Goldfunde aus Bulgarien. In Varnenskii nekropola i problemi halkolita, Symposium
Varnenskii nekropola i problemi halkolita, Symposium in- international, Varna, 19-23 avril 1976. Studia Praehis-
ternational, Varna, 19-23 avril 1976. Studia Praehistorica torica, 1-2, 170-181.
1-2, 27-45.
Todorova, H. 1978. The Eneolithic in Bulgaria. BAR, int.
Ivanov, I.S. 1978. Les fouilles archologiques de la n- ser. 49. Oxford: BAR.
cropole de Varna (1972-1975). Studia Praehistorica 1-2,
13-26.
Lichardus, J., Lichardus-Itten, M., Bailloud, G. and Cau-

vin, J. 1985. La protohistoire de lEurope. Nolithique et
Chalcolithique. Paris: PUF, Nouvelle Clio 1 bis.
Manolakakis, L. 2005. Les industries lithiques nolithi-

ques de Bulgarie. VML Internationale Archologie 88.
Manolakakis, L. 2006. Les trs longues lames de Varna,

quelle fonction? In J. Vaquer et J. Guilaine, La fin de
lAge de Pierre en Europe du sud, Actes de la table
ronde de lEHESS. Editions des Archives dEcologie
Prhistoriques, Toulouse, 2006. 5-23.
121
122
Marta Capote, Nuria Castaeda, Susana Consuegra, Cristina Criado, Pedro Daz-del-Ro
Flint mining in early neolithic Iberia: A preliminary report on Casa Montero (Madrid, Spain)
FLINT MINING IN EARLY NEOLITHIC IBERIA:

A PRELIMINARY REPORT ON CASA MONTERO
(MADRID, SPAIN)
Marta Capote, Nuria Castaeda, Susana Consuegra,

Cristina Criado, Pedro Daz-del-Ro
Abstract: The Early Neolithic (c. 5400-5100 cal BC) flint mine of Casa Montero was located in 2003 and partially exca-
vated throughout three field seasons. This paper describes the excavation strategies that have been applied both random
selection of shafts and aligned systematic sampling, and some of the preliminary results obtained from excavation and
post-excavation analyses. We conclude that the dimensions of the site and the characteristics of the shafts suggest short
term seasonal expeditions of small mining teams. This mining process produced use and probably exchange value, but
also a durable monumental landscape, erected as a result of a cumulative materialization of social labour.
Keywords: Flint mine, Neolithic, Iberia, operative chain, shafts, labour, monuments, landscape.
Introduction
Until the nineties most Iberian prehistoric flint mines had limitations of the small areas where digging took place do
been located in southern Spain through surface collec- not allow a precise functional interpretation of the earliest
tions. Of these, barely one had been explored (Ramos Mil- Neolithic layers, open-air settlements are mostly made up
ln 1997; Ramos Milln et al. 1997). All lacked absolute of small unimpressive clusters of pit-features. One of their
chronologies, and from what we now know, many were shared features is the limited amount of portable remains
actually mined in modern times for military purposes. In a recovered in early Neolithic layers. This is especially true
similar way, Neolithic studies have only recently commit- when quantifying flint tools, which seem to be extremely
ted to long-term regional research in areas other than the scarce, compared both to other contemporary evidence
well-known Levantine coast. elsewhere, or to the subsequent regional Copper Age
settlements (Daz-del-Ro 2004).
This is particularly the case for the Central Meseta, until
fairly recently an area almost devoid of Early Neolithic Under these circumstances, the discovery, in 2003, of the
evidence. Our present knowledge relies on a few recently Early Neolithic flint mine of Casa Montero in the centre
excavated sites, most known through preliminary reports of Iberia was a striking novelty on account of its location,
and only one completely published. There are two dif- early chronology, density of shafts, and the historical con-
ferent categories of sites: caves and open-air settlements text in which flint mining developed. The present paper is
(Bueno et al. 2002; Daz-del-Ro and Consuegra 1999; Es- an overview of the evidence recovered after three years
tremera 2003; Kunst and Rojo 1999; Rojo and Kunst 1999; of fieldwork, and a preliminary report on post-excavation
Rodrguez 2006). While the nature of the caves and the analyses in progress.
123
Fig. 1: Location of Casa Montero in Iberia and in relation to the Jarama basin.
1. The site and the digging campaigns
The flint mine of Casa Montero was discovered as a re- the excavation, characterized by an open-air Pleistocene
sult of the Archaeological Impact Assessment of Madrids stratigraphic sequence.
M-50 highway belt. The site is situated in the south-east of
Spains expanding capital city, in the centre of Iberia (Fig. The Neolithic evidence includes 3824 small vertical shafts,
1). Located on a river bluff, the position dominates one of one meter wide on average, and of up to nine meters deep,
the main regional river basins, the Jarama valley, where dependent on the variable depth and quality of the flint
some scattered Neolithic sites have been known to exist seams. The lithic record is exceptional, both in quantity
(Mercader et al 1989; Daz-del-Ro and Consuegra 1999). and quality, and includes all phases of the operative chain.
The results from three excavation seasons combined with Radiocarbon dating of two charcoal samples have afforded
an intensive surface survey with systematic sub-plowzone early Neolithic dates that suggest that the flint mine was
control pits suggest that the prehistoric mine may roughly exploited throughout the second half of the VI millennium
cover eight hectares. cal BC (Daz-del-Ro et al. 2006). To date, Casa Monte-
ro is, after Defensola, (Galiberti et al. 1997; 2001), the
Four chronological phases were documented in the 4 hec- second oldest Neolithic flint mine in Europe (Fig. 2).
tare open-area excavations undertaken at the site. These
offer evidence of Middle Pleistocene and Bronze Age Bronze Age pit structures are found in small percentages,
occupation, and Neolithic and Contemporary flint mining both in between the cluster of Neolithic shafts and in its
exploitations. The oldest phase was located in the north of contiguous area. None can be considered to be mining
124
Fig. 2: Calibrated radiocarbon dates of the earliest Neolithic flint mines in Europe: Defensola, Casa Montero, Krzemionki and Tomaszw.
shafts, but they share characteristics with all regional BA Copper Age enclosures (Daz-del-Ro 2004) or this Neo-
sites: the pits are mostly related to storage and waste dis- lithic flint mine. The distribution of features also shows a
posal. These shallow pits include a few crude throwaway pattern: a few structures on the previous site, while most
flint tools, potsherds, charcoal, some faunal remains, and cluster around in its immediate vicinity. Considering the
occasional individual burials. Many BA sites seem system- extremely irregular topography of the mine by the end of
atically to reoccupy certain previously significant sites, like the Neolithic, and the limited use of flint during the Bronze
125
Fig. 3: (A) Area documented during the first field season. (B) Area documented and sampled during the second and third field seasons. (C) Documented
area that has been preserved and protected.
Age, it is functionally unclear why scattered BA groups rez-Jimnez, 2005; Capote et al. 2006; Castaeda et al.
would occupy this particular spot. Nevertheless, the pres- 2008; Castaeda and Criado 2006; Consuegra et al. 2004;
ence of two individual burials suggests that at least some 2005; Daz-del-Ro et al. 2006; Daz-del-Ro et al. 2008;
ritual behaviour may have been involved. Whatever those Prez-Jimnez et al. 2005).
groups thought about previous flint mines or enclosures,
the pattern seems to suggest a strong interest in claiming That Casa Montero was the first Neolithic flint mine discov-
a connection with their -real or fictitious- ancestors, wher- ered and extensively dug in Iberia gave regional Heritage
ever they had left visible earthworks. Managers some arguments to push for a modification of
the highway belt some 60 meters to the west from its origi-
Finally, the surroundings of Casa Montero are known to- nal track, saving from destruction and permanently pro-
day as Cerro de las Canteras (Quarry Hill). Both the Neo- tecting the main concentration of prehistoric shafts. As a
lithic site and its surroundings were mined for flint to be result, during the second season (August 2004-February
used in threshing machines and possibly for gunflint. The 2005), the open-area excavation was extended 1.8 hectare
historical record suggests that they may have been under further west, and more than 500 new shafts were mapped.
exploitation until the mid-Nineteenth century (Consuegra This excavation led to reconsideration of our research
et al. 2004; Castaeda and Criado, 2006; Prado 1864). objectives, as key matters such as mining strategies, con-
temporaneousness of extractions or other issues could only
The site of Casa Montero was excavated during three field be potentially answered by recovering information from
seasons, from September 2003 to July 2006, under different clusters of shafts.
circumstances and with varying research aims (Fig. 3).
The first season (September 2003-March 2004) involved Consequently, during the third season (September 2005-
the excavation of 123 shafts, randomly selected from July 2006) the excavation strategy shifted from a focus on
the approximately 2500 shafts mapped in 2.4 hectares of individual mining shafts to groups of shafts. In order to
open-area excavation. The objective was a first assessment control potential horizontal variations in mining strategies
of the variability of prehistoric flint extracting methods, or chronology, we established an aligned systematic sam-
the geological structure of the site, and the absolute and pling. The approach involved the complete excavation of
relative dating of the site through the available evidence: all the shafts included in each one of the seventeen grid
charcoal remains and pottery fragments (Bustillo and P- squares of ten by ten meters systematically distributed
126
throughout the area that would ultimately disappear under 2. Geological context
the highway (Fig. 4).
Casa Montero is located in sedimentary rocks from the In-
For safety reasons, this sampling was accompanied by a termediate Unit of the Miocene in the Madrid basin (Brell
different excavation technique. All selected shafts were dug et al. 1985; Junco and Calvo 1983). The stratigraphic
down to two meters deep, followed by the complete extrac- column is composed of beds of clay, dolomite and silica
tion of the remaining geological layers with a backhoe. The rocks. Deep sections show the existence of four major sil-
grid squares were once again topographically located and icification episodes, each consisting of one or more silica
the excavation process repeated four times until most of the levels. The three upper levels are composed of opal and
deepest shafts had been completely dug out. Of all these, opaline chert, and were formed by silicification of Mag-
only three shafts were deeper than 10 meters and not com- nesian smectites (Prez-Jimnez et al. 2005; Bustillo and
pletely excavated. As a result, we obtained both detailed Prez-Jimnez 2005) (Fig. 5).
information on shafts and a precise three dimensional view
of the geological structure of the complete excavation area. Silica rocks from Casa Montero form nodules arranged in
This new excavation technique allowed for the recording discontinuous beds that may have some lateral continu-
of several narrow passages that connected groups of shafts, ity. These beds have been classified in four episodes of
a result that would have been impossible if we had fol- silicification (Bustillo and Prez-Jimnez 2005); they ap-
lowed the previous single shaft excavation procedure. Neo- pear deformed as a result of collapses of the underneath
lithic miners would have crawled through these passages evaporitic episodes. This deformation produced a depres-
following some horizontally disposed flint seams. sion in which most of the shafts are concentrated, and
Fig. 4: Aerial view of three of the 10x10 meter areas during their excavation.
127
may be particularly related to the horizontal depth vari-

ability of shafts throughout the site: as a general pattern,
shafts are deeper in the central area of the excavation.
The selection of this particular mining area was no coin-
cidence. The regional geology of Madrid is known for its
abundance in siliceous rocks; that is, the profusion of flint
at this site cannot be considered a uniquely distinctive fea-
ture. Although we cannot rule out the possibility of other
regional flint mines, Casa Monteros flint has a particular
genesis: it was formed from smectites and later underwent
an aging process. This involved a re-crystallization of opal
to form quartz in the inner part of nodules, developing a
fine-grained core. As a result, there are frequent nodules
with opaline outer parts and microcrystalline quartz in-
ner parts (Bustillo and Prez-Jimnez 2005). This process
gave the flint particularly suitable knapping qualities (P-
rez-Jimnez et al 2005; Bustillo and Prez-Jimnez, 2005;
Bustillo et al. 2008).
3. Raw materials
Casa Monteros lithic remains have been classified in two

groups of raw materials. First, rocks which were necessarily
obtained off-site, and used in the process of extraction,
quartering, and reduction of flint nodules. Secondly, silica
rocks, which were the main object of the mining activity.
Off-site rocks are quartzite, quartz, and sandstone peb-

bles, most probably obtained from the nearby Jarama river
bank, less than a kilometre away. Such rocks constitute a
minimal part of the overall assemblage.
The second group is composed of siliceous rocks: opals

and opaline cherts. Four silicification episodes have been
documented. Neolithic shafts cut through the three upper
episodes, all opaline. The main exploited levels are the
second and fourth, while the third is only exploited when-
ever it has good knapping qualities (Fig. 5).
We undertook a macroscopic and petrographic characteri-

zation of the raw materials represented in the archaeologi-
cal record, in order to analyze in depth the cultural and/
or physical characteristics that had determined either its
selection or refusal. In order to carry out this characteriza-
tion, a sample of 43 lithic remains was selected out of a
Fig. 5: Stratigraphical column documented at Casa Montero.
population of 5043 artefacts obtained from the first field
128
season. The macroscopic characterization has allowed us specific decision-making in respect to parameters such as
to distinguish seven types of raw materials. Subsequently, space, quality and quantity: excavation of the shafts, ex-
the petrographic study has defined the properties of the traction of raw materials, transformation of raw materials,
opaline cherts. and waste management. Each action has produced a par-
ticular kind of archaeological record.
Refits and chane opratoire analyses have provided evi-
dence indicating that when nodules have an opaline outer 4.1. Shaft excavation
part, this material is worked as a second phase of removal
of cortex, whereas the raw material chosen for lithic ex- In order to analyze this action, we rely on the following
ploitation is chert from the inner part of the nodules. evidence. First, the mining structures themselves and their
Precisely this inner part is the one reduced in the phases of spatial distribution. Second, the mining tools employed in
configuration and production. The reason for this selection the process of shaft digging and recovered in the archaeo-
is knapping quality, since opal is fragile and, consequently, logical record. Finally, the marks left by mining tools on
less suitable than chert as a blank. Moreover, the frailty shaft walls and the traces of preparation of structures to
of opals cause fractures, faults and errors in the process facilitate both access to shafts and evacuation of mine
of flaking. As a result, a large quantity of this raw mate- spoils.
rial inevitably has to be discarded. On the other hand, the
heterogeneity of nodules always involves additional diffi- Mining shafts offer little variability in respect to size and
culty for flint-knappers, as the response of each material to shape. They are mainly simple cylinder-like structures
the force of the blow will be unexpected and consequently on average one meter wide and up to nine or ten meters
difficult to control (Castaeda et al. 2008). (Fig. 6) deep. Their infillings show little differentiation, and few
archaeological remains other than an impressive amount
of flint. However, one of the sampling units dug during the
last field season revealed a cluster of five shafts with small
connecting galleries (Fig. 7). These were the result of side
workings dug to exploit the one meter deep seam. This
seam may have been abandoned when nearly exhausted or
when the costs and efforts to continue it exceeded those of
opening a new shaft. Occasionally, several of these small
connecting galleries are documented at different depths
in a single shaft. Although the system is more complex
and labour intensive than the straightforward excavation
of vertical structures, it is still extremely simple, given the
total scale of the mining activity.
Fig. 6: Flint refits from Casa Montero
4. The mine
The archaeological record of Casa Montero is the result

of sequential short term activities which took place over a
relatively short period of time. In order to tackle this typi-
cal problem of time scales, the information has been con-
textualized and sorted out distinguishing different working
processes. The labour process involved at least the follow-
ing succession of actions, each one of them preceded by Fig. 7: Side workings documented at one of the sampled areas.
129
As has been demonstrated for other Iberian mines (Cam- Two types can be identified among the raw material pro-
prub et al. 2003), Neolithic miners certainly had a good curement tools: striking tools and incisive tools. The first
knowledge of the geological structure of the area. Test group includes maces and big hammerstones; the second
pits at Casa Montero are an excellent testimony to this picks and wedges.
knowledge (Consuegra et al. 2004). The Neolithic miners
stopped excavating the shallow pits whenever they found Striking tools were heavy, with rounded forms. They were
green clay levels that are stratigraphically located under manufactured using raw materials obtained off-site, like
the opaline episodes. Furthermore, the depths of the shafts quartzite from the nearby terraces of the Jarama River. An
adapt to deformations resulting from siliceous episodes. example of this kind of tool is a mace manufactured from a
This geological know-how would have been the result of quartzite pebble, with prismatic shape and a quadrangular
a transmission of local mining knowledge from genera- and homogeneous section (Fig. 8). It measures 137x81x78
tion to generation: extraction and debitage methods, best mm and weights 1,400 grams, presenting two use areas
conditions for lateral excavations, safety measures, and on both ends, with traces of an intensive percussion that
so forth. Shafts were dug close together, none superim- provoked numerous accidental extractions. The mace has
posed on another, and with just enough distance so as to a medial zone with traces of crushing and abrasion, a con-
walk between them while avoiding wall collapses. These sequence of the elimination of protuberances and condi-
more than 4000 Neolithic shafts are probably the result of tioning for hafting. Maces may have been multifunctional,
several centuries of mining. The total mining intensity used for both the excavation of shafts and the fragmenta-
would have been about 13 shafts per year, considering a tion of big nodules (Consuegra et al. 2005).
time-span of 300 years for the whole period of activity.
Antler and bone tools for shaft digging are commonly

found in other European mines such as Jablines (Bostyn
and Lanchon 1992; 1997), Grimes Graves (Holgate 1995;
Topping 1997; Barber et al. 1999), Zele (Lech and Lech
1984; 1997) or Krzemionki (Migal and Salacinski 1997).
These tools may be picks, chisels, shovels or even rakes.
On the contrary, most of the tools related to mining ac-
tivities recovered so far from Casa Montero are made out
of stone. They are picks, maces or wedges, as is the case
for other mines like Ryckholt (Felder 1997) or Defensola
(Galiberti et al. 1997; 2001; di Lernia et al. 1995). The
very rare antler fragments recovered during the excavation
Fig. 8: An example of a flint pick and a quartzite mace from Casa Monte-
seem to relate more to small bone utensils such as awls ro.
or pressure flakers than to picks (Yravedra et al. 2008).
Nevertheless, the information recovered from several casts Incisive tools, on the other hand, were usually selected
of tool marks on shaft walls suggests the use of bone im- from discarded fragments that resulted from in situ knap-
plements in at least parts of the mining process. Both the ping activities. These fragments were chosen because of
analyses of these casts and of possible use wear on the their size, weight and pointed or angular morphology.
antler points will be decisive for their interpretation. Configuration is practically nonexistent in most of these
objects. When it does, it is simply performed in order to
Lithic mining tools have been classified in two groups, accord- facilitate hafting. They all show evident traces of percus-
ing to the kind of activity they permitted: raw material procure- sion and use.
ment tools -composed of shaft digging tools (picks, maces or
wedges)- and maintenance tools -mostly scrapers, denticulates Ten flint picks have been identified so far (Fig. 8), one of
and unmodified flakes- (Consuegra et al. 2004; 2005). them made out of silicified clay. The average dimensions
130
of these items are 103x53x27 mm. Their width and thick- They occasionally show evidence of intentional fractures, or
ness are normally more homogeneous than their length. the voids resulting from the withdrawal of whole nodules.
Working areas are usually very homogeneous, with a trian-
gular or more rarely a quadrangular or trapezoidal section; Resistance to fracture is greater when nodules are inserted
they usually are 15 mm thick. Most of them show longitu- in their matrix, and knapping becomes increasingly dif-
dinal and oblique striations on their active areas, and abra- ficult inside narrow shafts that do not allow the miner
sion on ridges. On most of them, microflakes and traces of to choose an adequate angle to strike. Depending on the
abrasion have been observed in their medial and proximal size of the nodule, and its position inside the shaft, nod-
ends. The light weight of these picks suggests that they ules would have been extracted whole or quartered and
may have not been used by themselves, but instead, em- extracted in large flakes. Those nodules small enough to
ployed as chisels in combination with maces (Consuegra be manipulated would have been directly removed from
et al. 2005). flint seams, while the bigger ones would have been frac-
tured with the aid of big hammerstones, maces and wedg-
We have classified denticulates, scrapers, endscrapers and es. Accordingly, there are few examples of exploitation
burins as maintenance tools. They were all expediently of complete nodules among the lithic production of Casa
manufactured from discarded flint fragments, and proba- Montero. As a rule, large flakes are extracted from nodules
bly used in activities such as the manufacture and repair of because they make more suitable core blanks (Consuegra
hafts, ladders or ropes. Non-modified flakes that resulted et al. 2005).
from knapping activities would also have been employed
for these maintenance activities. These would not have re- Tools used in raw material extraction can be divided into
quired any further modification, as the presence of flakes two groups: incisive implements such as wedges, and per-
with macroscopic use-wear traces -especially microflakes- cussion implements such as big hammerstones.
seems to suggest.
Some wedges have occasionally been recovered at the site.
Digging tools have occasionally left marks on shaft walls. These implements were either manufactured from dis-
Most of these are longitudinal marks with V or occasion- carded pieces resulting from flint knapping, or from frag-
ally U sections. They appear in groups of parallel marks ments of broken quarzite hammerstones, selecting those
with vertical or oblique disposition, although exceptionally with a particular morphology, size and weight, without
some shallow and circular marks have been documented. any specific configuration. Consequently, these wedges
Only occasionally do some shaft walls show certain modi- are improvised implements, as happens with all the tools
fications such as small depressions to facilitate access, or so far documented at Casa Montero. This kind of inci-
pairs of small opposite holes with circular section to sup- sive tool shows a thin distal end that increases in thick-
port some kind of pulleys. As the mean diameter of shafts ness towards its medial part and proximal end. Its working
does not allow for more than one miner at a time, any shaft area is an edge with highly variable length. This working
more than 2 meters deep would require a minimum of two edge has macroscopic marks on the edge, and striations
individuals to work it. The assistant stationed above would parallel to the main axis of the artefact. The proximal end
help with the process of waste disposal, hauling up the raw of these tools also display macroscopic use wear traces,
material, and handing down tools to the miner below. such as crushing and accidental extractions. These traces
are percussion marks resulting from the use of wedges as
4.2. Raw material extraction intermediate elements. The edge of the tool would have
been inserted into cracks or crevices in shaft walls, and the
This stage comprehends all the tasks performed in order to proximal end would have been struck with a mace in order
extract nodules from flint seams, quarter them, and raise to extract the nodules.
the nodules and fragments to the surface. Although shafts
do cut through flint seams, miners did not always exhaust The frequent difficulties occurring while quartering and
these, so some seams are still visible in the shaft walls. extracting nodules required the use of heavy hammer-
131
stones and, occasionally, the use of maces for shaft dig- Blade production is documented at the mine through
ging. These big hammerstones have been recovered either blade cores, crested blades, discarded blades and knap-
complete or more often as fragments. They are always ping accidents. Blade products of good quality do not
manufactured from quartzite, with a spherical or ellipti- appear, mainly because they were removed, probably
cal morphology, an approximate length of 163 mm and along with some configured blade cores. Neither used
an average weight of 1200 gr. They show many traces nor retouched blades have been recovered at the mine,
of hard percussion producing extensive accidental flakes that is, tool configuration and use are not represented in
and/or longitudinal fractures. the blade chane opratoire.
4.3. Raw material transformation. The exceptional lithic record of Casa Montero allows for
the recognition of different blade production schemes.
At the Casa Montero flint mine, the raw material was both The choice of one or another reduction strategy depends
procured and processed. This has resulted in a massive on the nature of the blank a complete nodule or a large
amount of recovered archaeological remains: lithic mate- flake- and on the blade needed. The lithic record analyzed
rials recovered from the three digging seasons weigh 65 from the first field season suggests that the preferred raw
T, and amount to more than one and a half million pieces. material for blade production was flint rather than opal.
This abundance of material creates extreme difficulties As documented through several refits, this preference re-
for its efficient management, processing, analysis and quired the elimination of all opal parts previous to the
comprehensive publishing. On the other hand, it offers configuration of the flaking surface (Consuegra et al.
the advantage of allowing the complete study and care- 2004; Castaeda et al. 2008).
ful reconstruction of the chane opratoire (Consuegra
et al. 2004; Castaeda and Criado 2006; Castaeda et al. When the blank is a whole nodule, the systems developed
2008). are mainly prismatic, beginning with the preparation of a
crested blade. Flake blanks are more suitable and versatile
A preliminary analysis of lithic remains from the first than entire nodules. These large flakes allow for different
field season led us to understand the main aim of flint reduction strategies. So far, three reduction schemes for
exploitation: the production of blades. There is also a blade production have been documented, two in volume
secondary, although important, flake production, iden- and one in surface. Each will produce a particular kind
tified with its own independent chane opratoire. This of blade: either more elongated and thin or shorter and
kind of production is peculiar to Casa Montero, espe- thicker (Castaeda and Criado 2006; Daz-del-Ro et al.
cially when compared to other European Neolithic flint 2006).
mines that mostly aimed at the production of axe rough-
outs, such as Grimes Graves (Barber et al. 1999), Jablines Flake production has been identified thanks to the presence
(Bostyn and Lanchon 1997) or Defensola (Galiberti et al. of predetermined and predetermining flakes and their
1997). Examples of Neolithic flint mines aiming at blade corresponding cores. Flakes are obtained through bifa-
production, such as Tomaszw (Schild 1995), are some- cial exploitation systems, either hierarchical or non-hier-
how less common in Western Europe. archical: both discoidal and levallois schemes have been
identified (Consuegra et al. 2004).
Evidence from Neolithic contexts suggests that blades
played both a functional and ritual role. Blades have been Hammerstones used for knapping activities were collected
documented in early Neolithic individual burials from and selected among materials from the Jarama river bank.
Central Iberia, such as La Lmpara (Rojo and Kunst 1999), They are quartzite, quartz and sandstone pebbles that can
although they later become a generalized funerary offering be sorted in two different volumetric groups: medium
in megaliths (e.g. Rojo et al. 2005; Delibes et al. 1993). and small, fitting an average hand.
Not infrequently, blades were specifically manufactured
for burials, showing no evidence of wear for use.
132
4.4. Waste Management will be obtained in order to define the potential techno-
logical and/or compositional variability between wares.
The soil extracted during the excavation of shafts was either These may inform us about functional differences between
dumped into other nearby shafts along with the remaining wares, or differences in the provenance of pottery produc-
waste from flaking, or left aside and finally dumped back tions, perhaps related to the social conditions of accessibil-
in. This of course depends on whether miners opened more ity to the mine.
than one shaft at a time. It seems reasonable to suggest
that the second option was preferred: we lack evidence On the other hand, there is a small assemblage of bone
for natural sediments at the base of any excavated shaft, tools, most of them corresponding to bone and antler
and layers that can be interpreted as wall collapses are shafts, and finished rings that were produced and aban-
rare. In any case, it seems that shafts were filled almost doned on site (Yravedra et al. 2008). This evidence sug-
immediately after being excavated. Shaft in-fills show a gests that the manufacture of bone and antler rings was
repetitive pattern where density of lithic finds increases somehow required while labouring at the mine. Although
with depth. This behaviour would keep working areas free the function of these rings and their relation with flint
from spoils and flaking residues, and, given the proxim- mining are yet to be found, it seems that these artefacts
ity of shafts, the systematic backfilling may have been the -typical of Iberian Levantine Early Neolithic sites (Pascual
most practical safety measure. Benito 1996)- should be considered rather as functional
than ornamental, as previously thought. Whatever their
5. Evidence of other activities function, they must relate to an activity performed in both
mining and non-mining contexts, and they are one of the
Miners at Casa Montero understandably performed other best chronological markers of Casa Montero.
activities. Among the archaeological evidence recovered
are pottery fragments, bone industry, faunal remains and Finally, ochre impregnated on lithic elements and bones,
ochre fragments. has been documented in considerable quantity, as well as
a large 2 kg portion of pigment. Ochre has usually been
The few fragments of pottery recovered at Casa Montero related to both ritual and functional contexts: it was used
are in form and decoration characteristic of central Iberias in funerary monuments, to process different raw materials,
Early Neolithic, similar -if not identical- to those recovered and in the production of elements such as hafts. Never-
at Cueva de la Vaquera (Estremera 2003; Consuegra et al. theless, there is so far no unambiguous evidence for Neo-
2004). Frequently they are fragments of impressed wares lithic ritual activities at Casa Montero, as opposed to other
(Fig. 9), occasionally decorated with red dip-coatings, Iberian (Bosch and Strada 1994) or European mines (e.g.
and their shapes suggest mostly middle-sized containers. Barber et al. 1999; Topping 1997). However, mining itself
Pottery will be analyzed for phytoliths, and thin-sections may have been a ritualized practice.
6. Concluding remarks
Some preliminary conclusions can be drawn from Casa

Montero, especially when compared to European Neolithic
flint mines in general.
The first observation bears on the scale of mining activity.

This Iberian mine is noticeably small in extension, com-
pared to many other western European flint mines. This
is not so strange, if we consider the most probable low
population densities of the Madrid region during the late
Fig. 9: Two impressed ware fragments recovered at Casa Montero. Sixth millennium BC, and the 300 year life-span of the
133
mining activity suggested by radiocarbon dates (see Fig. The second observation refers to the main purpose of flint
2). These similarities are less obvious when comparing exploitations throughout Neolithic Europe. While many
Casa Montero with the flint mines of Defensola (Galiberti European mines are associated with the manufacture of
et al. 1997; 2001), its contemporary central Mediterranean flint axes, some with outstanding aesthetic qualities (e.g.
counterpart. Although they are both associated with the Whittle 1995), the flint of Casa Montero lacking all
first introduction of a production economy in their respec- evident aesthetic quality was dedicated to the produc-
tive regions, differences in the size and complexity of la- tion of blades. In fact, there is no evidence in Iberia of the
bour and mining techniques seem obvious. use of flint for the manufacture of polished axes, mainly
because of the wide variety of available hard stones such as
The shafts are also considerably smaller than in most sillimanite, ofite or amphibolite (e.g. Lillios 1997; Orozco
flint mines, with the 117 cm mean diameter of shafts at 2000).
Casa Montero (Fig. 10). They allow for no more than one
1
miner working inside at a time, all of which suggests that Prehistoric flint mines such as Casa Montero are particu-
the minimum Neolithic working team may not have been larly suitable archaeological sites for the analysis of how
more than two individuals. However, we should consider labour was organized and deployed by the first Neolithic
the possibility of some labour coordination throughout societies. It is difficult to think of a site that materializes
the whole mining and knapping procedure, as seems to be labour in a more direct and manifest form. There is abso-
the case in most known ethnographic cases (Burton 1984; lutely no evidence of human activity other than flint min-
Stout 2002). Whatever the size of the mining teams, the ing. It was all production, with practically no consumption.
width of the shafts seems to result from the combination of Nevertheless, there is an aspect that should not be underes-
a limited availability of labour force and practical reasons timated: the construction of monumental spaces and their
related to the geological structure of the area, where col- use in the process of social reproduction. Unquestion-
lapses are common when clays are exposed. ably, one of the most relevant features of the Neolithic is
a massive increase in the quantity of durable materiality
(Hodder 2005: 131). The mining process produced both
use and probably exchange value, but also involved
the progressive creation of a collective space. The first
regional monumental space was erected as a result of the
cumulative materialization of social labour. Generations of
Neolithic miners created a durable monumental landscape
while reinforcing the mechanisms of social reproduction
involved in the transmission of communal rights over
resources.
Marta Capote
Instituto de Historia
Centro de Ciencias Humanas y Sociales, CSIC
C/ Albasanz, 26-28, 28037, Madrid
e-mail: mcapote@ih.csic.es
Fig. 10: A deep mining shaft from Casa Montero compared to scale with Nuria Castaeda
Grimes Graves Pit 1.
Centro de Ciencias Humanas y Sociales, CSIC
C/ Albasanz, 26-28, 28037, Madrid
1 Compare, for example, with the 232 cm mean diameter of the excava-
ted shafts at Jablines (Bostyn and Lanchon 1992). e-mail: ncastanyeda@ih.csic.es
134
Susana Consuegra Bustillo, M. A., Castaeda, N., Capote, M., Consuegra, S.,
Instituto de Historia Criado, C., Daz-del-Ro, P., Orozco, T., Prez-Jimnez,
Centro de Ciencias Humanas y Sociales, CSIC J. L. and Terradas, X. 2008: Is the macroscopic classifi-
C/ Albasanz, 26-28, 28037, Madrid cation of flint useful ? A petroarchaelogical analysis and
e-mail: sconsuegra@ih.csic.es characterization of flint raw materials from the Iberian
Neolithic mine of Casa Montero. Archaeometry, DOI:
Cristina Criado 10.1111/j.1475-4754.2008.00403.x.
Centro de Ciencias Humanas y Sociales, CSIC Bustillo, M. A. and Prez-Jimnez, J. L. 2005. Caracters-
C/ Albasanz, 26-28, 28037, Madrid ticas diferenciales y gnesis de los niveles silceos ex-
e-mail: cctorija@ih.csic.es plotados en el yacimiento arqueolgico de Casa Montero
(Viclvaro, Madrid). Geogaceta 38, 243-246.
Pedro Daz-del-Ro
Instituto de Historia Bueno Ramrez, P., Barroso Bermejo, R., Balbn Behrman,
Centro de Ciencias Humanas y Sociales, CSIC R., Campo Martn, M., Etxeberra Gabilondo, F., Gonzlez
C/ Albasanz, 26-28, 28037, Madrid Martn, A., Herrasti Erlogorri, L., Treserras, J. J., Lpez
e-mail: diazdelrio@ih.csic.es Garca, P., Lpez Sez, J. A., Matamala, J. C. and Snchez,
B. 2002. reas habitacionales y funerarias en el Neoltico
de la cuenca interior del Tajo: La provincia de Toledo. Tra-
REFERENCES bajos de Prehistoria 59 (2), 65-79.
Barber, M., Field, D. and Topping, P. 1999. The Neolithic Camprub, A., Melgarejo, J-C., Proenza, J.A., Costa, F.,
Flint Mines of England. Swindon: English Heritage. Bosch, J., Estrada, A., Borell, F., Yushkin, N.P. and Andre-
ichev, V.L. 2003. Mining and geological knowledge during
Bosch, J. and Strada, E. 1994. La Venus de Gav (Barce- the Neolithic: a geological study on the variscite mines at
lona). Una aportacin fundamental para el estudio de la Gav, Catalonia. Episodes 26(4), 295-301.
religin neoltica en el suroeste europeo. Trabajos de Pre-
historia 51(2), 149-158. Capote, M., Castaeda, N., Consuegra, S., Criado, C.,
Daz-del-Ro, P., Bustillo, M. A. and Prez-Jimnez, J. L.
Bostyn, F. and Lanchon, Y. (eds.) 1992. Jablines. Le Haut 2006. Casa Montero, la mina de slex ms antigua de la
Chteau (Seine-et-Marne). Une minire de silex au No- Pennsula Ibrica. Tierra y Tecnologa 29, 42-50.
lithique. Documents dArchologie Franaise, 35. Paris:
Editions de la Maison des Sciences de LHomme. Castaeda, N., Capote, M., Criado, C., Consuegra, S.,
Daz-del-Ro, P., Terradas, X. and Orozco, T. 2008. Las ca-
Bostyn, F. and Lanchon, Y. 1997. The Neolithic Flint Mine denas operativas lticas de la mina de slex de Casa Monte-
at Jablines, Le Haut Chteau (Seine-et-Marne). In A. Ra- ro (Madrid). IV Congreso del Neoltico en la Pennsula
mos-Milln and M.A. Bustillo (eds.), Siliceous Rocks and Ibrica (Alicante, 27-29 de Noviembre de 2006). Volume
Culture, Granada : Universidad de Granada,271-291. 2, Alicante: Museo Arqueolgico de Alicante, 231-234.
Brell, J. M., Doval, M. and Carams, M. 1985. Clay min- Castaeda, N. and Criado, C. 2006. La industria ltica de
eral distribution in the Evaporitic Miocene Sediments of Casa Montero (Viclvaro, Madrid): resultados prelimin-
the Tajo Basin, Spain. Minealogica et Petrographica Acta ares. In N. Ferreira and H. Verissimo (eds.). Do Epipale-
29-A, 267-276. oltico ao Calcoltico na Pennsula Ibrica. Actas do IV
Congresso de Arqueologia Peninsular (Faro, 14 a 19 Se-
Burton, J. 1984. Quarrying in a Tribal Society. World Ar- tembro de 2004), Faro, 229-234.
chaeology16 (2), 234-247.
135
Consuegra, S., Castaeda, N. and Gallego, M. M. 2005. Felder, P. J. S. 1997. Excavations of Prehistoric Flint
Explotacin minera de slex neoltica de Casa Montero Mines at Ryckholt-St. Geertruid (Netherlands), Grimes
(Viclvaro, Madrid). Avance de los resultados de las exca- Graves and Harrow Hill (United Kingdom). In A. Ramos-
vaciones. In O. Puche and M. Ayarzagena (eds.), Minera Milln and M.A. Bustillo (eds), Siliceous Rocks and Cul-
y metalurgia histrica en el sudoeste Europe, 45-51. Ma- ture, 143-150. Granada: Universidad de Granada.
drid.
Galiberti, A., Di Lernia, S., Florentino, G. and Guariscio,
Consuegra, S. Gallego, M. M. and Castaeda, N. 2004. M. 1997. New Data on The Neolithic Flint Mine of Defen-
Minera neoltica de slex de Casa Montero (Viclvaro, sola, Vieste (Italy). In A. Ramos-Milln and M.A. Bustillo
Madrid). Trabajos de Prehistoria 61(2), 127-140. (eds.), Siliceous Rocks and Culture, 211-222. Granada:
Universidad de Granada.
Delibes, G., Rojo, M. and Represa, I. 1993. Dlmenes de
La Lora: Burgos. Junta de Castilla y Len. Consejera de Galiberti, A., Sivilli, S. and Tarantini, M. 2001. La miniera
Cultura y Turismo. neolitica della Defensola (Vieste-Foggia): Lo statu delle
ricerche. Origini XXIII, 85-110.
Daz-del-Ro, P. 2004. Copper Age ditched enclosures in Cen-
tral Iberia. Oxford Journal of Archaeology 23 (2), 107-121. Hodder, I. 2005. The spatio-temporal organization of the
early town at atalhyk. In D. Bailey, A. Whittle and
Daz-del-Ro, P. and Consuegra, S. 1999. Primeras eviden- V. Cummings (eds), (Un)settling the Neolithic, 126-139.
cias de estructuras de habitacin y almacenaje neolticas Oxford: Oxbow.
en el entorno de la Campia madrilea: el yacimiento de
La Deseada (Rivas-Vaciamadrid, Madrid). In J. Bernabeu Holgate, R. 1995. Neolithic flint mining in Britain. Ar-
Aubn and T. Orozco Khler (eds) II Congrs del Neoltic a chaeologia Polona 33, 133-161.
la Pennsula Ibrica, 251-257. Valncia: Saguntum Extra.
Junco, F. and Calvo, J. P. 1983. Cuenca de Madrid. In W.
Daz-del-Ro, P., Consuegra, S., Capote, M., Castaeda, N., Gibbons and T. Moreno (eds), Geologa de Espaa. IGME,
Criado, C., Orozco, T. and Terradas, X. 2008. Estructura, Tomo II, 534-543.
contexto y cronologa de la mina de slex de Casa Monte-
ro (Madrid). IV Congreso del Neoltico en la Pennsula Kunst, M. and Rojo, M. 1999. El Valle de Ambrona: un
Ibrica (Alicante, 27-29 de noviembre de 2006). Volume ejemplo de la primera colonizacin Neoltica de las tierras
1, Alicante: Museo Arqueolgico de Alicante, 200-207. del Interior Peninsular. In J. Bernabeu Aubn and T. Oro-
zco Khler (eds), II Congrs del Neoltic a la Pennsula
Daz-del-Ro, P., Consuegra, S., Castaeda, N., Capote, Ibrica, 259-270. Valncia: Saguntum Extra.
M., Criado, C., Bustillo, M. A. and Prez-Jimnez, J. L.
2006. The earliest flint mine in Iberia. Antiquity 80 (307). Lech, H. and Lech, J. 1984. The Prehistoric Flint Mine at
http://antiquity.ac.uk/ProjGall/diazdelrio/index.html Wierzbica Zele: A Case Study from Poland. World Ar-
chaeology 16(2), 186-203.
Di Lernia, S., Fiorentino, G., Galiberti, A. and Basili, R.
1995. The Early Neolithic mine of Defensola A (I 18): Lech, H. and Lech, J. 1997. Flint Mining among Bronze
flint exploitation in the Gargano area. Archaeologia Polo- Age Communities: A Case Study from Central Poland.
na 33, 119-132. In R. Schild and Z. Sulgostowska (eds), Man and Flint.
Proceedings of the VIIth International Flint Symposium.
Estremera, M. S. 2003. Primeros agricultores y ganaderos en Warszawa-Ostrowiec witokrzyski, 91-98. Warszawa:
la Meseta Norte: el Neoltico de la Cueva de la Vaquera (Tor- Institute of Archaeology and Ethnology Polish Academy
reiglesias, Segovia). Arqueologa en Castilla y Len. Conse- of Sciences.
jera de Educacin y Cultura. Junta de Castilla y Len.
136
Lillios, K.T. 1997. Amphibolite tools of the Portuguese Rodrguez Gonzlez, D. 2006. Historia de la investigacin
Copper Age (3000-2000 BC): a geoarchaeological study y estado de la cuestin del Neoltico en Castilla-La Man-
of prehistoric economics and symbolism. Geoarchaeology cha: una visin de conjunto. In N. Ferreira y H. Verissi-
12(2), 137-163. mo (eds), Do Epipaleoltico ao Calcoltico na Pennsula
Ibrica. Actas do IV Congresso de Arqueologia Peninsular
Mercader, J., Corts, A.F. and Garca, M.E. 1989. Materi- (Faro, 14 a 19 Setembro de 2004), 217-228.
ales neolticos en el valle del Jarama (Arganda, Madrid).
Trabajos de Prehistoria 46, 255-260. Rojo, M. and Kunst, M. 1999. La Lmpara y la Pea de
La Abuela. Propuesta secuencial del Neoltico Interior en
Migal, W. and Salacinski, S. 1997. Studies at Krzemionki el mbito funerario. In J. Bernabeu Aubn and T. Orozco
during the last decade. In R. Schild and Z. Sulgostowska, Khler (eds), II Congrs del Neoltic a la Pennsula Ibri-
Man and Flint. Proceedings of the VIIth International Flint ca, Valncia: Saguntum Extra, 503-512.
Symposium. Warszawa-Ostrowiec witokrzyski, 103-
108. Warszawa: Institute of Archaeology and Ethnology Rojo, M., Kunst, M., Garrido, R., Garca, I. and Morn,
Polish Academy of Sciences. G. 2005. Un desafo a la eternidad: tumbas monumentales
del valle del Ambrona, Soria. Memorias de Arqueologa
Orozco, T. 2000. Aprovisionamiento e Intercambio: Anli- en Castilla y Len, 14, Consejera de Cultura y Turismo.
sis petrolgico del utillaje pulimentado en la Prehistoria Valladolid : Junta de Castilla y Len.
Reciente del Pais Valenciano (Espaa). B. A. R. Int. Ser.
S867, Oxford: BAR. Schild, R. 1995. Tomaszw, Radom Province. Archaeolo-
gia Polona, 33, 455-465.
Pascual-Benito, J. L. 1996. Los anillos neolticos de la
Pennsula Ibria, Congrs del Nolitic a la Pennsula Ibri- Stout, D. 2002. Skill and Cognition in Stone Tool Produc-
ca (Gav-Bellaterra). Rubricatum I, 279-289. tion. An Ethnographic Case Study from Irian Jaya. Current
Anthropology 43(5), 693-722.
Prez-Jimnez, J.L., Alonso-Zarza, A.M. and Bustillo,
M.A. 2005. Sedimentologa y diagnesis de las secuencias Topping, P. 1997. Structured deposition, symbolism, and
lacustres someras en el rea del yacimiento neoltico de the English flint mines. In R. Schild and Z. Sulgostowska
Casa Montero (Viclvaro, Madrid). Geo-Temas 8, 77-80. (eds), Man and Flint. Proceedings of the VIIth Internation-
al Flint Symposium. Warszawa-Ostrowiec witokrzyski,
Prado, C. de, 1864. Descripcin fsica y geolgica de la Warszawa: Institute of Archaeology and Ethnology Polish
provincia de Madrid. Madrid: Imprenta Nacional. Academy of Sciences,127-131.
Ramos-Milln, A. 1997. La Venta. A prehistoric Flint Mine Yravedra, J., Maicas, R., Daz-del-Ro, P. and Consuegra,
in a Tribal Society (Iberian South East).I n R. Schild and S. 2008. Anillos para una mina. Industria sea y fauna de
Z. Sulgostowska (eds), Man and Flint. Proceedings of the la mina de slex neoltica de Casa Montero (Madrid). IV
VIIth International Flint Symposium. Warszawa-Ostrow- Congreso del Neoltico en la Pennsula Ibrica (Alicante,
iec witokrzyski, Warszawa: Institute of Archaeology 27-29 de noviembre de 2006). Volume 2, Alicante: Museo
and Ethnology Polish Academy of Sciences, 117-121. Arqueolgico de Alicante, 240-247.
Ramos-Milln, A., Osuna-Vargas, M.M., Tapia-Espinosa, A., Whittle, A. 1995. Gifts from the earth: symbolic dimen-
Pena-Gonzlez, B. and Aznar-Prez, J.C. 1997. Archaeologi- sions of the use and production of Neolithic flint and stone
cal Research in the La Venta Flint Mine (Iberian Southeast). axes. Archaeologia Polona 33, 247-259.
In A. Ramos-Milln and M.A. Bustillo (eds), Siliceous Rocks
and Culture, Granada: Universidad de Granada, 245-270.
137
138
Yoann Gauvry: Intensive extraction of non-metallic minerals during the later prehistory in the northern half of Europe
INTENSIVE EXTRACTION OF NON-METALLIC

MINERALS DURING THE early protoHISTORY
IN THE NORTHERN HALF OF EUROPE
Yoann Gauvry
Abstract: The main aim of this work is, firstly, to determine the state of knowledges about mineral exploitation available
in archaeological literature, in order to study, secondly, problematics elaborated on a geographically and chronologi-
cally large scale. The techniques for acquiring non-metallic mineral raw materials, which include every degrees between
ponctual surface gathering untill deep and organized mines, are full of informations about the people who employed them ;
it is interesting to understand their evolution during the different periods of Neolithic, Chalcothic and early Metal Ages,
and to see if they obey global and common rules in the northern half of Europe.
Keywords: Mines, quarries, stone, flint, raw material extraction, Neolithic, Chalcolithic, Europe.
Introduction
This paper will present the main issues and results of a chronological. This is essential in order to grasp their full
general study on the contribution of mining archaeology to historical and cultural potential.
our understanding of later Prehistory, especially during the
Neolithic and Chalcolithic periods. It concerns the north- The geographical scope of this review covers the northern
ern half of Europe, and is conceived in the light of previ- half of Europe, from the dawn of the Neolithic until the
ous well known studies on a global scale, such as those beginning of the Bronze Age (Fig. 1). All the published
of Clark and Pigott 1933, Jahn 1960, Schmid 1973 and data regarding the archaeology of the principal stone
1974, Lech 1981a-b Each of these studies has shown mines and quarries have been collected; a corpus of a hun-
the importance of replacing the data on stone exploita- dred and ten exploitation sites has been brought together,
tion within a broader framework, both geographical and which constitutes the basis of this overall study (Fig. 2).
Early Protohistory
6000
Early Neolithic
Middle Neolithic
DK S 5000
Late Neolithic
IRL
GB NL
4000 Fig. 1: Geographical and chronological
B
D
PL Early Chalcolithic frame of the study
CZ
F
SK
A
H
CH
3000 Middle Chalcolithic
Late Chalcolithic
2000
Early Bronze Age
B.C.
139
Fig. 2: Localisation of the 110 sites treated in the study
a b c d
g h
Opencast exploitation Slope exploitation Underground exploitation - mining

a : surface gathering e : slope quarrying g : underground extension
b : surface quarrying f : slope tunneling h : underground network
c : shallow quarrying
d : deep quarrying
Fig. 3: the main exploitation technics, illustrated by flint extraction.
140
Among the several hundred references identified during Thus, the ultimate objective of this work can be summa-
the past 160 years, these particular sites have been selected rized by these three questions:
on the basis of the good, or at least acceptable, quality of -When and how did intensive mining of stone appear in
information concerning them. The corpus includes sites Europe?
of all sizes, depths, densities; they have been the object -How did it develop?
of many excavations, which has made it possible to rec- -What does it teach us about early protohistory periods?
ognize various types of exploitation techniques: opencast
mining and underground galleries, simple and complex 1. Chronological review
sites, rudimentary or well organized production techniques
(Fig. 3) 1.1. Definition
Studying stone mines and quarries requires that one takes In order to treat a global question of early protohistory, it is
into account several specific difficulties, which occur to all necessary to adopt large contemporary horizons to divide
levels of interpretation. The complexity of the formation this era. Here, the early protohistory subdivision and ter-
of exploitation structures and of their fill, as well as their minology defined in the Protohistoire de lEurope (Lichar-
frequent lack of artefactual material, makes understanding dus et al. 1985) appears to us to be the most appropriate
and dating them difficult. Moreover, it is rarely possible to choice. Therefore, unless it is specifically mentioned, all
establish clear associations between particular extraction the chronological stages of early protohistory (late Neo-
techniques and precisely dated cultural complexes. Taking lithic, middle Chalcolithic, etc.) must be understood in this
these difficulties into account, we can nonetheless propose pan European meaning.
three principal stages in such a study:
-First: to compare these sites with one another, and to con- 1.2. Neolithic
struct a typology of the extraction methods and techniques ;
-Second: to deduce how technical procedures were devel- There is little positive evidence of Neolithic exploitation
oped in response to socio-economical demands; structures (Fig. 4). In many mines and quarries, some Neo-
-Third: to place these data along a chronological axis, and lithic material and absolute dates have been identified; but
translate them into chrono-cultural interpretations. they are rarely clearly related to exploitation structures.
Exploitaition structures of La Defensola (Italy)

(Di Lierna et al. 1995: 121-122)
Fig. 4 : exploitation sites known and supposed for the middle and late Neolithic.
141
early Funnel Beaker

culture
Windmill Hill late Lengyel

culture culture
post-Rssen
underground cultures
opencast
Chassen -
Michelsberg
Deep quarrying in Saspow (Poland) (Lech in Babel et al. 1999: 618)
Opencast and underground structures in Jablines (France)

Deep quarrying in Arnhofen (Germany) (Gayck 2000: 270)
(Bostyn and Lanchon 1992: 213)
Underground extension in Hov (Denmark) (Becker and Weisgerber 1999: 463) Underground network in Harrow Hill (Great-Britain) (Shepherd et al. 1999: 521)
Fig. 5 : exploitation sites of early Chalcolithic, and details.
142
When material has been found on the surface of a site, it is 1.3. Early Chalcolithic
not necessarily possible to establish a direct relation with
the exploitation structures. The artefacts might be related A great many of the best known and most spectacular
to a previous exploitation phase, in which entirely different mines and quarries are attributed to the early Chalcolithic
techniques were employed. Moreover, most of the raw ma- (Fig. 5). However it is important to recall that throughout
terials employed during the Neolithic epoch were also ex- the Chalcolithic, these great mines and quarries coexisted
tracted during the following era (Chalcolithic), and the pri- with small and short-lived exploitations, which are at-
mary structures masked or destroyed by the later extraction tested at every phase of the periods under review. We will
phases. For example, we know that Jurassic Krakow flint conduct this review from East to West, following the major
was habitually used by Neolithic societies, but the main cultural groups.
exploitation sites must be attributed to Chalcolithic times,
according to several radiocarbon datings (Lech 1981a, Late Lengyel culture
48) and to cultural characteristics as well (Dzieduszycka-
Machnikowa and Lech 1976: 167-168; Lech 1981b: 275). Mineral resources attract populations, who settle in their
vicinity in spite of soils most unfavourable to agriculture.
By way of illustration, here are a few interesting cases. The This fact is evident at the site of Szentgl (Transdanubia,
ochre exploitation structures of Bad Sulzburg (Black For- Hungary), where mineral resources attracted populations
est, Germany), excavated in the late 1980s, provided sev- who settled nearby, despite poor agricultural terrain (Bir
eral radiocarbon calibrated datings around 5000 BC (Gold- and Regenye 2003: 59-60).
enberg et al. 2003: 185). However, the similarity between
these exploitation structures and tools and those of the jas- This phenomenon also occurs in regard to the Jurassic
per exploitation at Kleinkems, very close to Bad Sulzburg, Krakow flint, which played a major role ahead during
and dated 1000 years later, invites caution about attributing Neolithic times and during the Chalcolithic, when im-
this site to the Neolithic. portant deposits of this material attracted new groups of
people to previously uninhabited areas (Lech 1981b: 275).
In the reliably dated Italian mine of La Defensola, flint was Important exploitation sites developed, such as at Saspw
extracted in large galleries during early and middle Neo- (wojw. Krakw, Poland), where an intensive extraction by
lithic. This is a particularly interesting site. The exploita- deep quarrying was set up to obtain nodules in sub-original
tion was carried out through galleries 60 cm high on aver- positions in a layer of clay several meters deep. Raw ma-
age, which formed a network of more than 6000 m on the terial was not only obtained, but was also partially trans-
level of the higher bed. The waste products were contained formed on the site. The production steps with the highest
behind low dry stone walls, carefully arranged to preserve a risk of failure, and creating the largest quantity of waste,
clean and safe passage. Ceramics and many extraction tools was carried out on the site in order to limit the amount
were found on the exploitation floor; this material, in addi- of non-useful material to be transported (Dzieduszycka-
tion to several radiocarbon datings, indicates a Neolithic Machnikowa and Lech 1976).
context (Galiberti et al. 1997: 220).
On the Vienne-Mauer (Austria) site, the extraction of ra-
La Defensola, although outside of our geographical re- diolarite in much degraded chalk was carried out in shafts
search area, raises very exciting questions about the old- which reached a depth of 6 to 10 m, interconnected by
est indisputably underground mines known in Europe. The horizontal galleries (Ruttkay 1970). A feature of particular
same questions can also be asked about the Spanish flint interest here is a burial site with seven individuals, men,
mine of Casa Montero, dated around 5400-5200 cal BC, women and children, probably linked to the mining activi-
which shows all the characteristics of the later mines (see ties.
capote et al., this volume). These two exceptions highlight
the need to extend our study to the southern half of Europe,
and eventually to open it to the Mediterranean context.
143
Post-Rssen cultures In the present state of knowledge, the record depth level
has been observed at Spiennes (Hainaut, Belgium), with
A number of Jurahornstein deposits are found throughout shafts 16 m deep at Camp--Cayaux. There is a certain
Bavaria, on both sides of the Danube. This raw material contrast between the different exploitation areas, which for
was used from the middle Neolithic on, but it is difficult to the most part were in activity in the Michelsberg culture
specify the exploitation techniques employed at that time. (Collet 2004: 131).
We can mention the large shallow pits of Flintsbach-Hardt,
and the deep shafts of Arnhofen (Bavaria, Germany) in As for the Chassen, many important sites are known. The
which silexite nodules were extracted from a thick layer Jablines site (Seine-et-Marne, France) shows a great variety
of surface clay. Unfortunately these structures are impre- of structures, from shallow pits to shafts 7 m deep with an
cisely dated, as the raw materials were probably extracted extremely rational organisation so as to best cover the de-
at different periods in the same place. Several elements posit (Bostyn and Lanchon 1992). We should also mention
seem to indicate that the principal phases of mining took the Villemaur-sur-Vanne Les Orlets site (Aube, France),
place during the post-Rssen and Mnchshfen cultures where four flint beds were exploited in structures very close
(Eisele et al. 2003: 61). together, sometimes overlaying one another: a clear illustra-
tion of the geographical concentration of vestiges from dif-
On the Kleinkems (Baden-Wrttenberg, Germany) site, ferent exploitation phases (De Labriffe et al. 1995: 110).
the slope quarrying and slope mining structures and the
tools were very similar to those found in the ochre exploi- North-western Europe cultures
tation in the Black Forest, which is, however, 1000 years
older according to radiocarbon dating. Fire was used to We note that the appearance of intensive extraction con-
fragment the chalk, which was then ground up with heavy cerns the whole continent, even the peripheral regions of
quartzite hammers in order to free up the beds of jasper. north-western Europe, which had not yet been reached by
Unfortunately, we dont know how the material was used the Neolithic way of life.
in the region. This exploitation was fairly intensive. The
site was clearly dated by radiocarbon samples and by ce- For example, underground mining is a feature of the early
ramic shards present in two tombs clearly attributed to the Funnel Beaker culture (FBC) in southern Scandinavia, as
Munzingen group (Schmid 1952: 12; Engel and Sigmund can be seen at Kvarnby (Malm district, Suede), or at Hov
2005). (Thisted district, Denmark). The intensive exploitation of
flint beds here can reach down to 8 m, with wooden struc-
Michelsberg / Chassen tures detectable in the shaft fills. In England, the under-
ground flint exploitation was used from the start, cf. the
The Michelsberg and Chassen cultures are well docu- Windmill Hill culture. This is clearly visible on the South
mented, and include some of the most spectacular mines. Downs sites, particularly at Harrow Hill and Long Down
(West Sussex, Great Britain).
The Rijckholt-St. Geertruid (Limburg, Netherlands) site,
with its extensive underground networks, illustrates ex- 1.4. Middle Chalcolithic
tremely rational and efficient exploitation techniques for a
deposit whose depth varies between 2 and 12 m (Felder et Thereafter, the main tendencies observed in early Chal-
al. 1998). The transformation of the raw material, essen- colithic seem to continue during the middle Chalcolithic,
tially during the stages carrying the highest risk of failure even though we have less information concerning that pe-
was partly carried out on site: roughouts of axes and and riod (Fig. 6).
pre-core shaping, debitage of blades, before transport away
from the site to finish off the production. These products
played an important regional role during the Michelsberg
culture.
144
late Funnel Beaker Culture /

Global Amphorae Culture
Seine-Oise-Marne /
Wartberg
underground
opencast Altheim
Exploitation structures and tools in Krzemionki (Poland)

(Migal 1997: 101 ; Borkowski 1997: 333)
Shallow quarrying in Scernfelder Forst (Germany) (Gayck 2000: 284) Great slope quarrying in Lousberg (Germany) (Weiner and Weisgerber 1999: 105)
Fig. 6 : exploitation sites of the middle Chalcolithic, and details.
145
Late FBC ple Valkenburg flint, extracted nearby. Another source of

growing importance was the flint mined in the Lousberg
In Poland, we observe a few changes compared to the pre- hill (Ldkr. Aachen, Germany), through a technique of
ceding period: after the predominance of the Jurassic Kra- large-scale slope levelling which almost razed the 4 up-
kow flint and chocolate flint, we note the important role per meters of the hill. Finally, we note the continuation of
played by the flint extracted at Swieciechow (wojw. Tar- mining activities at Spiennes during the Seine-Oise-Marne
nobrzeg, Poland). This high quality raw material is found culture (SOM), principally in the Petit-Spiennes aera,
at less than a meter below the surface, and spreads out over through underground galleries not so deep as earlier (Col-
a large area (570 km) in the form of axes and blades (Bal- let et al. 2004: 157).
cer 1976: 190). It hides the appearance of the Krzemionki
striped flint site at the beginning of the middle Chalcol- 1.5. Late Chalcolithic
ithic, first exploited in opencast structures.
The late Chalcolithic era is not so well documented (Fig. 7).
It is important to note that at the Gorzsw site (wojw. However, several global tendencies may be observed, specifi-
Gorzsw Wielkopolski, Poland), flint nodules were cally in the peripheral regions of Europe.
gathered in the gravel of a small river, and used directly
in a local settlement (Bbel et al. 1999: 624). This type of In most of the sites, modest, opencast and shallow pit-
exploitation is very light, leaving few traces behind, and mining was common, with the raw material locally dis-
must have been quite common over the whole of early pro- seminated and used. Some of these had been intensively
tohistory, satisfying small-scale demands other than those mined during earlier periods, chocolate flint for instance,
that led to developing mines of more consequence. but during the late Chalcolithic these products are less fre-
quently found.
At the end of middle Chalcolithic, during the Globular
Amphorae culture (GAC), we observe the acme of striped We do however know of some intensive exploitation sites.
flint exploitation at Krzemionki (wojw. Tarnobrzeg, Among these the sites of Krasnoye Selo and Karpautsy
Poland) (Migal 1997b: 324). There, the variety of the de- (Belarus), and especially Grimes Graves (Norfolk, Great
posit layout entailed a large spectrum of extraction tech- Britain), where we find spectacular underground networks.
niques, in order to obtain a raw material of high quality, These few examples of intensive underground mineral ex-
favoured for its attributes both mechanical and aesthetic. traction are geographically near major mining sites of an-
cient and middle Chalcolithic.
Altheim culture
Finally, some more modest exploitation sites display
Widespread use of Jurahornstein continued in Bavaria, specific characteristics: they are linked to the produc-
but the exploitation techniques of this period are poorly tion of new types of artefacts of non-functional character.
documented and dated. There are indications that raw ma- Take, for example, the Danish and Swedish flint daggers
terial was scarce and intensively extracted at sites such as (Becker 1959: 88), they were made from raw material ob-
Baiersdorf and Schernfleder Forst (Bavaria, Germany). tained through shallow and simple structures such as the
ones of Aalborg (Denmark), and even collected on coastal
Northwestern Europe beaches such as Fornaes (Denmark). The production of
Grand Pressigny flint artefacts in France obeyed similar
During the middle Chalcolithic, particularly in the Wart- rules: one notes the same contrast between simple and
berg culture, Lixhe-Lanaye flint extracted at Rijckholt-St. modest extraction techniques and the production of elegant
Geertruid is still commonly found, obtained in the same 40 cm long blades requiring very sophisticated individual
structures as during the previous period. However, other know-how (Pelegrin 2002), and which seem intended
raw materials played a more important role, for exam- more for ceremonial than for practical use.
146
Underground network in Krasnoye Selo (Belarus) (Gurina 2000: 86)
underground
opencast
Underground structures in Grimes Graves (Great Britain) (Felder 1999: 121)
Opencast structures supposed for the extraction of Grand Pressigny flint (France) (Dijkman et al. 1999: 484)
Fig. 7 : exploitation sites of late Chalcolithic, and details.
147
1.6. Early Bronze Age and after 2. Interpretation
Stone extraction continued to decline during the early 2.1. The Neolithic-Chalcolithic transition
Bronze Age, when metal production came to play the ma-
jor role (Fig. 8). However, in some areas distant from the Through technological, petro-archaeological and func-
main axis of metal diffusion, some kinds of stone extrac- tional analyses, we know that Neolithic stone economies
tion survived. For example, several opencast and under- were in various ways complex and geographically exten-
ground extraction sites have been attributed to the culture sive, involving a large spectrum of strategies to obtain
of Mierzanowice, in Poland. raw material. Except in a couple of cases, exploitation
techniques used then are not well known; we may sup-
In several mining sites of Central and Western Europe, pose that pit-mines were the rule, and that these extraction
such as Rijckholt-St. Geertruid and Grimes Graves, some activities were regularly kept up, but not with the same
Iron Age occupation is attested both by ceramics and by intensity as during the later periods. We may assume that a
absolute dating techniques. Such artefacts seem episodic Neolithic substratum largely determined subsequent de-
and slight, compared to the extensive and regular extrac- velopment. It already included most of the features which
tion of the previous periods. characterised, to various degrees, the stone economics of
underground
opencast
Fig. 8 : exploitation sites and traces of Metal Ages.
148
the Chalcolithic: complementarity, interdependence, ex- early Chalcolithic, throughout the whole northern half of
tension over sometimes very large geographical areas, pro- Europe. The main outcome was the expansion of the range
duction and exchange of artefacts of a more complex nature of mining techniques; an intensive and technologically
than simply functional tools (Lech 1979: 45). These facts sophisticated production system with a capacity for deep
suggest intense and large-scale contacts already developed mineral extraction spread quickly through the north-west-
among Neolithic communities and cultural groups. ern margins of Europe (Scandinavia and British Islands).
We also conclude that the appearance of new demands,
There was no real innovation as regarding extraction tech- involving new technological responses, coexisted with
niques probably employed during the Neolithic and those economic systems already established during the Neolithic
in use from the beginning of the Chalcolithic. Neolithic and which persisted for some time.
people knew where to find raw materials, and how to dig
deep to extract them, as the water wells found in central 2.2. Mineral exploitation during the Chalcolithic
Europe also attest. However, there was probably not the
economic demand to exploit these raw materials as in- During the early Chalcolithic, we observe the largest spec-
tensively as was done during the Chalcolithic. The new trum of mining techniques for the whole early early pro-
mining techniques appeared during this epoch are there- tohistory. The most extensive and elaborate techniques,
fore linked to a growth of demand, not a change in its such as deep quarrying of thick deposits and systematic
nature, which was similar to the previous era and obeyed underground networks, were generally adopted all over
the same basic rules (Lech 1991: 569). the northern half of Europe. Thus raw materials, for the
Accordingly, we shall not argue for a break between Neo- large-scale production of standardised artefacts, mainly
lithic and Chalcolithic traditions of mineral exploitation, axes and blades, were extensively sought after, intensively
as could have happened in the case of a real technologi- exploited and sometimes widely diffused.
cal innovation. This conclusion refutes the hypothesis
that the origins of intensive mining techniques should be These products suggest a new notion of specialisation. Dur-
associated with the ore-mining occurring with the first ing the early Chalcotlithic, the deepest and most intensive
metal production. extraction techniques involved specialised populations,
with considerable mining engineering skills, acquired
However, it would be wrong to conclude that there was a collectively and over generations of part-time miners. The
simple linear evolution between these two eras. At the be- chanes opratoires of both extraction and production led,
ginning of the Chalcolithic, mineral exploitation not only with some exceptions, to a generally efficient organization
displayed a new diversity and intensity, but also involved of standardised and easily communicable tasks, not requir-
specialisation to a greater extent: communities now settled ing highly individual skills.
near the sources of minerals, in agriculturally poor areas
previously uninhabited; the different phases of work were It is not easy to define the characteristics of these mining
efficiently distributed in space and in accordance with in- populations. Some evidence, such as the presence of
dividual abilities Even though these phenomena may abundant mining waste products in the surrounding set-
have been present to some extent during the Neolithic, tlements, suggests that they may have been established as
during the Chalcolithic they took on so much importance village communities. However, the overall lack of compel-
that they generated a real mining engineering expertise in ling evidence incites us to characterise them with notions
some populations. of milieu technique, or even groupe technique, as defined
in the works of A. Leroi-Gourhan (1971: 321 ; 1973: 348).
Thus we see a clear transformation in regard to mineral In any case, these populations did not necessarily carry on
exploitation between the end of the Neolithic and the long-extended mining activities, but may have done so in
beginning of the Chalcolithic. Because of the extended, repeated, regular and perhaps annual periods.
interconnected economic systems of the Neolithic, these
changes happened almost simultaneously, during the
149
We shall therefore not compare these activities, as is Conclusion

sometimes done in archaeological literature (see for
example Gardiner 1990: 122), to large-scale and infrequent The conclusions driving from this study do not end
collective works, such as the erection of megaliths or the discussions of this topic, far from it. Rather, they are meant
construction of causewayed enclosure, which probably to offer a general scheme of interpretation, which needs to
required the authority of an individual master-builder. be nuanced and no doubt corrected by new research into
On the contrary, the collective experience involved in mineral exploitation, particularly through dialogue with
intensive mineral exploitation would have been progres- the other disciplines concerned with lithic economies. In
sively elaborated, through a repeated practice included in all events, these general observations are meant to show
an ordinary and probably annual economic organization. how the study of extraction techniques can contribute to
defining and better understanding the main periods of
The main tendencies of the early Chalcolithic econ- early early protohistory, which really deserves to be more
omy persisted widely thereafter. During the middle widely and more precisely studied
Chalcolithic, new importance seems to have been given to
the aesthetic properties of raw materials. This evolution,
which may sometimes be linked to the production of pres- Yoann Gauvry
tige pieces with a clear symbolic value, did not directly UMR 7041 ArScAn, quipe Protohistoire europenne
involve changes in the nature and the scale of the mineral 5 bvd des Allis, 91720 MAISSE
exploitation, but played an increasing role in the choice of e-mail: yoann.gauvry@wanadoo.fr
the raw materials.
During the middle and late Chalcolithic, a progressive nar- References

rowing of the spectrum of exploitation techniques can be
observed. Except in some notable cases, both the intensity Bbel, J., Balcer, B., Chmielewska, M., Dagnan-Ginter,
and the complexity of these productions declined. The role A., Ginter, A., Lech, J. and Schild, R. 1999. Katalog der
of underground galleries clearly decreased; at the same Bergwerke Poland. In Weisgerber G., Slotta R., Weiner
time, the opencast and slope extractions, requiring a sim- J. (dir) : 5000 Jahre Feuersteinbergbau - Die Suche nach
pler know-how, gained in relative importance. Moreover, dem Steinzeit, Exposition au Deutschen Bergbau-Museum
the products of mines exploited only for a short time, with Bochum, du 24 oct. 1980 au 31 jan. 1981, ditions du
rough and ready techniques, were not so much obscured Deutschen Bergbau-Museum Bochum, n77, 3me dition
by the predominance of standardised artefacts from major amliore, largie et actualise, 578-627.
production sites as had earlier been the case.
Balcer, B.1976. Position and stratigraphy of flint deposits,
During the late Chalcolithic, we observe almost exclu- development of exploitation and importance of the Swie-
sively simple opencast extraction techniques, sometimes ciechow flint in Prehistory. A.A.C. t. XVI, 179-199.
linked with the production of at least partly non-functional
items, such as the Danish and Swedish flint daggers and Becker, C. J. 1959. Flint mining in Neolithic Denmark.
the long blades of Grand-Pressigny. These productions re- Antiquity 33, 130, 87-92.
quire high individual skills and experience, developed by
each professional knapper personally, contrasting with Becker, C.J., Weisgerber, G. 1999. Katalog der Berg-
the collective nature of the know-how involved by inten- werke Dnemark. In Weisgerber G., Slotta R., Weiner
sive and complex mining techniques. This kind of profes- J. (dir) : 5000 Jahre Feuersteinbergbau - Die Suche nach
sionalism, based on prestige craft production and appear- dem Steinzeit, Exposition au Deutschen Bergbau-Museum
ing at the very end of the Chalcolithic, may constitute the Bochum, du 24 oct. 1980 au 31 jan. 1981, ditions du
first step towards the emergence of skilled artisans during Deutschen Bergbau-Museum Bochum, n77, 3me dition
the Bronze Age. amliore, largie et actualise, 456-473.
150
Bir, K.T. and Regenye, J. 2003. Exploitation regions and ber G., Slotta R., Weiner J. (dir) : 5000 Jahre Feuerstein-
workshop complexes in the Bakony Mountains, Hungary. bergbau - Die Suche nach dem Steinzeit, Exposition au
In Stllner Th., Krlin G., Steffens G., Cierny J. (dir.) : Deutschen Bergbau-Museum Bochum, du 24 oct. 1980 au
Man and Mining - Mensch und Bergbau. Studies in hon- 31 jan. 1981, ditions du Deutschen Bergbau-Museum Bo-
our of Gerd Weisgerber on occasion of his 65th birthday, chum, n77, 3me dition amliore, largie et actualise,
Der Anschnitt Beiheft 16, Bochum, 55-63. 474-509.
Borkowski, W. 1997. Exploitation field in Krzemionki. Di Lierna, S., Fiorentino, G., Galiberti, A. and Basili, R. 1995.
System of deposit utilization. In Siliceous Rocks and The early Neolithic mine of La Defensola A (I18): flint ex-
Culture Ramos-Millan et M.A. Bustillo Eds. VI Interna- ploitation in the Gargano area. In Lech J. (dir.) : Special Theme:
tional Flint Symposium, Madrid, 1-4 octobre 1991, Gra- Flint Mining, Actes du 7 symposium international du silex,
nada, 327-335. Varsovie Ostrowiec Swietokrzyski 4-8 septembre 1995, Ar-
chaeologia Polona, vol. 33: 1995, Institute of Archaeology and
Bostyn, F. and Lanchon, Y. 1992. Jablines, le Haut Ch- Ethnology, Polish Academy of Sciences, Warsaw, 119-132.
teau (Seine et Marne); une minire de silex au Nolithique.
Document dArchologie Franaise 35. Paris: Maison des Dzieduszycka-Machnikowa, A. and Lech, J. 1976. Ne-
Sciences de lHomme. olityczne zespoy pracowniane z kopalni krzemienia w
Sspowie [The neolithic workshop assemblages from the
M. Capote, N. Castaeda, S. Consuegra, C. Criado, P. Daz- flint mine of Sspow], Polskie Badania Archeologiczne t.
del-Ro ce volume. Flint mining in early Neolithic Iberia: a 19. Polska Akademia Nauk - Instytut Historii Kultury Ma-
preliminary report on Casa Montero (Madrid, Spain). terialnej.
Clark, G. and Piggott, S. 1993. The age of the British flint Eisele, Kl., Rind, M. M. and Sorcan, B. 2003. Ausgrabungen
mines, Antiquity 7, 166-183. und Dokumentationen im neolithischen Hornsteinbergwerk
von Abensberg-Arnhofen 2000 bis 2002, Rind, 42-51.
Collet, H. 2004. Les mines nolithiques de Spiennes : tat
des connaissances et perspectives de recherche, actes du Engel, F. and Siegmund, Fr. 2005. Radiocarbon dating
XIV congrs UISPP, section 9 : Le Nolithique au Proche of the Neolithic flint mine at Kleinkems (near Efringen-
Orient et en Europe, 129-133. B.A.R. International Series Kirchen, District Lrrach, Baden-Wrttemberg, Ger-
1303, Oxford: BAR. many), Antiquity 79, 306. (website http://antiquity.ac.uk/
projgall/siegmund/index.html).
Collet, H., Collette, O. and Woodbury, M. 2004. Indices
dextraction et de taille du silex datant du Nolithique r- Felder, P. J. 1999. Feuersteinbergbau in Ryckholt-St-Geer-
cent dans la carrire CBR Harmignies. Note prliminai- truid und Grimes Graves - ein Vergleich. In Weisgerber
re. Notae Praehistoricae 24, Studia Praehistorica Belgica, G., Slotta R., Weiner J. (dir): 5000 Jahre Feuersteinberg-
151-158. bau - Die Suche nach dem Steinzeit, Exposition au Deut-
schen Bergbau-Museum Bochum, du 24 oct. 1980 au 31
De Labriffe, P.-A., Augereau, A., Sidra, I. and Ferdouel, jan. 1981, ditions du Deutschen Bergbau-Museum Bo-
F. 1995. Villemaur-sur-Vanne Les Orlets (Aube), qua- chum, n77, 3me dition amliore, largie et actualise,
trime et dernire minire de lautoroute A5; rsultats pre- 120-123.
liminaries. 19 colloque interrgional sur le Nolithique,
Amiens 1992, Revue Archologique de Picardie, n sp- Felder, P. J., Rademakers, P., Cor, M. and De Grooth, M.
cial 9, 105-119. E. Th. 1998. Excavations of prehistoric flint mines at Ri-
jckholt-St. Geertruid (Limburg, The Netherlands), by the
Dijkman, W., Guillaume, Ch., Affolter, J. and Roden, Ch. Prehistoric Flint Mines Working Group of the Dutch
1999. Katalog der Bergwerke Frankreich. In Weisger- Geological Society, Limburg Section, Archologische
151
Berichte 12. Bonn: Deutschen Gesellshaft fr Ur- und chardus J. (dir.): Die Kupferzeit als historische Epoche,
Frhgeschichte. Symposium Saarbrcken und Otzenhausen 6.-13.11.1988,
Saarbrcker Beitrge zur Altertumskunde, Band 55, Bonn,
Galiberti, A., Di Lierna, S., Fiorentino, G. and Guarascio, 2 vol., 557-574.
M. 1997. New data on the Neolithic mine of Defensola,
Vieste (Italy). In Ramos-Millan et M.A. Bustillo (eds). Leroi-Gourhan, A. 1971. Evolution et techniques : lhom-
Siliceous Rocks and Culture VI International Flint Sym- me et la matire. Paris: Albin Michel.
posium, Madrid, 1-4 octobre 1991, Granada, 211-222.
Leroi-Gourhan, A.1973. Evolution et techniques : milieu
Gardiner, J. 1990. Flint procurement and Neolithic axe et techniques. Paris: Albin Michel.
production on the South Downs: a re-assessment. Oxford
Journal of Archaeology 9(2), 119-141. Lichardus, J., Lichardus-Itten, M., Bailloud, G. and Cau-
vin, J. 1985. La Protohistoire de lEurope, le Nolithique et
Gayck, S. 2000. Urgeschichtlicher Silexbergbau in Europa le Chalcolithique. Paris: Presses Universitaires de France,
- Eine kritische Analyse zum gegenwrtigen Forschung- Nouvelle Clio.
stand. Weissbach: Beitrge zur Ur- und Frhgeschichte
Mitteleuropas 15. Migal, W. 1997a. Selected aspects of specialization in
mining and flint knapping. In Schild R., Sulgostowska Z.
Goldenberg, G., Maass, A., Steffens, G. and Steuer, H. (eds) : Man and flint, actes du 7 symposium international
2003. Hematite mining during the Linear Ceramics Cul- du silex, Varsovie Ostrowiec Swietokrzyski 4-8 septem-
ture in the area of the Black Forest, South West Germany. bre 1995, Institute of Archaeology and Ethnology Polish
In Stllner Th., Krlin G., Steffens G., Cierny J. (dir.): Academy of Sciences, Warszawa, 99-101.
Man and Mining - Mensch und Bergbau. Studies in hon-
our of Gerd Weisgerber on occasion of his 65th birthday, Migal, W. 1997b. Reconstruction of the flint extraction
Der Anschnitt Beiheft 16, Bochum, 179-186. system in Krzemionki. In Ramos-Millan et M.A. Bustillo
(eds). Siliceous Rocks and Culture VI International Flint
Gurina, N. N. 2000. Prhistorische Feuersteinbergwerke in Symposium, Madrid, 1-4 octobre 1991, Granada, 315-
der ehemaligen UdSSR. Der Anschnitt Beiheft 12. Bochum. 325.
Lech J. 1979. Flint mining among the early farming com- Pelegrin, J. 2002. La production des grandes lames de silex
munities of Poland. Neerlandse Geologische Vereiniging, du Grand-Pressigny. In Guilaine J. (dir.) : Matriaux, pro-
39-45. ductions, circulations du Nolithique lAge du Bronze,
sminaire du Collge de France, Collection des Hespri-
Lech, J. 1981a. Flint mining among the early farming des, Errance, Paris, 131-148.
communities of central Europe, Przegld Archeologiczny,
(Wrocaw) 28, 5-55. Ruttkay, E. 1970. Dasjungsteinzeitliche Hornsteinberg-
werk mit Bestattung von der Antonshhe bei Mauer (Wien
Lech, J. 1981b. Grnitwo krzemienia Spoecznoci wczes- 23). Die Ausgrabungen Josef Bayers in den Jahren 1929-
norolniczych na Wyynie Krakowskiej, koniec VI tysiclecia 1930. Mitteilungen der Anthropologischen Gesellschaft
1 poowa IV tysiclecia p.n.e. [Flint mining among the 100, Wien, 70-83.
early farming communities in the Cracow upland (end
of VIth millenium to mid-IVth millenium B.C.)], Polska Shepherd, R., Sieveking, G. de G., Weisgerber, G. and Ro-
Akademia Nauk - Instytut Historii Kultury Materialnej. den Ch. 1999. Katalog der Bergwerke Grossbritannien.
In Weisgerber G., Slotta R., Weiner J. (dir) : 5000 Jahre
Lech, J. 1991. The Neolithic-Eneolithic transition in Feuersteinbergbau - Die Suche nach dem Steinzeit, Ex-
prehistoric mining and siliceous rock distribution. In Li- position au Deutschen Bergbau-Museum Bochum, du 24
152
oct. 1980 au 31 jan. 1981, ditions du Deutschen Bergbau-

Museum Bochum, n77, 3me dition amliore, largie
et actualise, 510-543.
Schmid, E.1952. Jungsteinzeitliches Jaspis-Bergwerk am

Isteiner Klotz. Der Anschnitt 5/1952, 4. 4-14. Bochum.
Schmid, E.1973. Die Reviere urgeschichtlichen Silexberg-

baus in Europa. Der Anschnitt 4/1973, 25, 12-15. Bo-
chum.
Schmid, E.1974. Die Reviere urgeschichtlichen Silexberg-

baus in Europa - Teil IV: Die Silex-gewinnung in Nord-
frankreich. Der Anschnitt 3/1974, 18-22. Bochum.
Weiner, J. and Weisgerber, G. 1999. Die Ausgrabun-

gen des jungsteinzeitlichen Feuersteinbergwerks Lous-
bergww in Aachen 1978-1980 (D3), Ein Vorbericht. In
Weisgerber G., Slotta R., Weiner J. (dir) : 5000 Jahre
Feuersteinbergbau - Die Suche nach dem Steinzeit,
Exposition au Deutschen Bergbau-Museum Bochum, du 24
oct. 1980 au 31 jan. 1981, ditions du Deutschen Bergbau-
Museum Bochum, n77, 3me dition amliore, largie et
actualise, 92-119.
153
154
Paul Wheeler: Ideology and Influences behind the Neolithic flint mines of the Southern Britain
Ideology and Influences behind the

Neolithic flint mines of the Southern Britain
Paul Wheeler
Abstract: This paper presents the results of part of a research project which aims to evaluate the influences and evolution
of the early Neolithic in Southern Britain. The following discourse endeavours to understand the development of the early
Neolithic flint mining tradition, with particular reference to the Sussex mines of southern Britain.
The chronologically contemporary flint mines of Western Europe are analysed, in order to determine if common practices
occur amongst the mines of Europe and those that first appeared in Britain and thus uncover any potential continental
influences on the early Neolithic of Southern Britain. The research looks beyond superficial similarities, and looks to
identify commonalities in the ideology and the attitude associated with the mines and the mining process. Do we find
similar events taking place in the mines? Are technical problems dealt with in similar ways? and how does the approach
towards flint manifest itself? The common practices identified are then analysed in order to determine if an ideological
bond can be found throughout the flint mining process of Western Europe. Ideas are proposed about what we can infer
from communities who are displaying a similar set of cultural choices across this part of Europe.
Keywords: Flint mine, western Europe, Neolithic.
Introduction local communities had the desire to obtain flint from deep
within the landscape. Why did this change take place?
In Southern Britain one of the first truly monumental struc- Perhaps there was a need for a greater quantity or purer
tures to be created within the early Neolithic were the flint quality of flint. Or perhaps an altogether different ideology
mines of the Sussex downs. The landscape begins to be and attitude towards the procurement of flint was created.
peppered with these deeply intrusive, exotic and complex The question thus arises as to whether the idea and ability
subterranean structures, and so begins one of the earliest to mine flint was an indigenous innovation or could such
truly monumental endeavours of the local communities of concepts have been introduced from the continent.
Britain. The technological and ideological concepts behind Thus the main objective regarding the analysis of flint
the creation of the mines would presumably have been a mines is to critically evaluate the evidence for a continen-
revolutionary concept, with far reaching implications on tal influence for their establishment in the British land-
the politics and economics of the societies involved. scape in the early Neolithic.
In order to establish any potential influences on the devel-
Thus the initial appearance of the flint mines prompts two opment of the mines, this paper looks beyond a morpho-
observations. Firstly, a change in the mindset and mental- logical and technological analysis, upon which variation
ity of the populous must have taken place. Previously local and commonalities can be explained via environmental de-
communities would have collected flint from river beds terminism, and looks at the cultural choices undertaken by
or from outcrops (Barber 2001: 22-23), however, now each community. If there was a continental influence upon
155
the flint mines of Sussex, one would expect to identify a Across the channel we find the French mines of Bretteville-
series of common cultural practices, which are inherently le-Rabet (Desloges 1986) and Jablines (Bostyn and Lanchon
associated within flint mining traditions. 1992 ; 1997) are being exploited within a similar period, as
are the Belgic mines of Jandrain Jandrenouille (Hubert 1974;
1. Background to the research Weigsberger 1980: 429-431) and the mines at Spiennes (Hu-
bert 1978; Weigsberger et al. 1980: 418-421). The Dutch mine
There where three main criteria for the selection of com- of Rijckholt-St-Geertruid (De Grooth 1998; Bosch 1979;
parative flint mine sites throughout Europe. Felder et al. 1998; Felder 1981) is also in use during a this
Firstly the sites had to fall into a chronologically contem- time (Fig. 2).
porary period as the Sussex Mines. Thus this study will The chronology of the sites is based on the latest and earliest
examine the sites that were in operation in the last few recorded calibrated dates obtained for each site under study.
centuries of the 5th millennium BC and for the first half of The date ranges indicate the mines were in operation at ap-
the 4th millennium BC. The chronology for the mine sites proximately the same time, the data indicates an average com-
are based on the earliest recorded calibrated dates. mon exploitation period of approximately 3800BC (Fig. 2).
Secondly the mines had to lie within the pre determined
study region, which consists of Britain, Belgium, France, 2. What are the Common Practices?
The Netherlands and western Germany (Fig. 1). Thirdly,
only mines containing a large enough excavated and re- A number of common cultural choices can be explored, the
corded sample range were analysed, in order to have a jus- concept of mining for flint, the flint seam exploited, the
tifiably comparable dataset. The British mines under study products manufactured at the mine and finally an analysis
consist of the Sussex mines of Church Hill (Barber et al. of the various types of deposits within mine shafts. The
1999; Russell 2001: 85-158), Harrow Hill (Curwen and first commonality between the Sussex mines and those on
Curwen 1926; Holleyman 1937), Cissbury (Russell 2001: the European continent is the construction and concept of
170-192; Topping 2005: 63-93) and Blackpatch (Pull flint mining itself.
1932; Russell 2001: 24-84). The flint mines themselves are the construct of a totally
Fig. 1: Western European Flint mining sites and those under study.
156
Fig. 2: Early Neolithic Flint Mining chronology for Western Europe.
alien technological ability. Nothing on this scale had previ- Thus the overall concept and cultural choice to mine for
ously been created within southern Britain. Thus the initial flint can be identified as a common practice across the
occurrence of flint mines equates not only to a new concept continent and thus equate to potential cultural influence.
but a new technological skill amongst the local commu-
nities who now not only posses the engineering ability to 3. The Choice of flint seam to be
construct these great subterranean monuments but also the exploited
desire to intrusively exploit the earth for this raw material.
But how did the indigenous populations obtain such skills The process of flint mining is largely determined by envi-
and evolve the desire to mine for flint? ronmental factors. Firstly, you can only mine in a geologi-
Looking across to the European mainland we find that cal landscape that contains flint and of a reasonable qual-
the mining communities at Spiennes, Rijckholt and those ity, so communities had little choice over location.
other mines in the study region, are harnessing and ex- Secondly the size of a shaft is determined by the structural
ploiting similar methods in the extraction of flint (Felder integrity of the underlying geology. This also affects the
et al. 1998; Hubert 1978). Whether in the form of small tools used to dig shafts.
pits, quarries or big, deep shafts and subterranean galler- Thus the first quantifiable culturally determined choice,
ies, the concept of digging deep into the earth is abun- that miners appear to have made is, which flint seam
dantly evident, as are the technological and engineering should be exploited?
skills used to sustain such a vast monumental complex Examination of the shafts evaluated to date have identified
(Pull 1932: 26). that the majority of the miners throughout this period in
157
Europe were not exploiting the first or second seams of Only at Blackpatch was the first seam encountered ex-
flint they encountered during the process of extraction, but ploited, but even here, as amongst the other Sussex shafts,
were exploiting the deeper seams (Fig. 4), often bypassing miners would dig down up to 15ft to access the desired ma-
the more easily accessible flint above (Curwen and Cur- terial (Pull 1932: 25). Such cultural choices of the Sussex
wen 1926; Felder et al. 1998: 39; Pull 1932: 25). miners are witnessed amongst the mines of the European
mainland. At Spiennes the miners would often dig through
The evidence from the Sussex Downs shows that at Church numerous seams to exploit the lowest safely accessible
Hill the first 3 seams of flint were by-passed before ex- seams, reaching depths of up to 16m (Fig. 4). The flint at
ploitation took place (Russell 2001: 85-158). At Cissbury this level was up to 30cm thick, quite different from the
the 3rd and 4th seams were those exploited (Russell 2001: seams above (Hubert 1978). The miners at Rijckholt would
179-192), and at Harrow Hill it was mainly the 3rd seam also by-pass the first easily exploitable seams, looking to
to be exploited, with the first seams dug straight through utilise the specific lanaye band of flint (Fig. 3), which com-
(Fig. 3) (Curwen and Curwen, 1926). monly occurred at seam 10 (Felder et al. 1998, 12).
Fig. 3: Average flint seam exploited in the mine sites under study.
Fig. 4: Maximum depht of shaft within the mine sites of under study.
158
At Jandrain-Jandrenouille and Jablines, although 1st and At the larger mines of Spiennes and Rijckholt, again we
2nd seams were exploited (Fig. 3), the required flint con- find axe manufacture integrally associated with these
tinued to be difficult to access; often these mines reached mines, however blade manufacture appears to be an im-
depths of 5m and 6m respectively (Hubert 1974; Bostyn portant part in their later phases (Hubert 1978; Weigs-
and Lanchon 1992: 58-87). berger 1980: 418-421; De Grooth 1998; Bosch 1979;
Felder et al. 1998; Felder 1981).
The evidence from across Europe shows that mining com- Axe manufacture appears to be integrally linked to flint
munities were making a cultural choice to either exploit mining, particularly in the early Neolithic (Topping
particular seams of flint, or they were choosing to exploit 1997). All the mines have a strong association with the
seams which were more difficult and dangerous to access production of flint axes. If mine shafts are only to be used
(Barber 2001: 23). for the production of axes this may explain some of the
Figure 4 shows the comparative depths the deepest mines choices identified so far. Perhaps it is this association that
at each site would reach. But what determined this choice? determines which flint seams are exploited?
In regards to cases where specific seams are exploited, it Firstly perhaps only particular seams are of a fine enough
appears that implicit values may be attached to these par- quality for the production of axes, for example the La-
ticular seams. naye layer at Rijckholt, the thick seams at Spiennes or
Where miners are choosing to dig deep into the earth, even the utilised layer at Church Hill. But why were the
and bypassing other exploitable seams, we may be wit- lesser quality seams not also exploited for more com-
nessing episodes on a par with those witnessed amongst monplace products? Perhaps the mine shafts were solely
stone axe procurement sites, where it is the difficulty of the prestige of axes, and thus the lesser seams could not
obtaining the raw material that creates the social values be utilised for the production of secondary (less prestig-
imbedded within the objects of production (Topping ious) products.
2005: 83-84). Thus perhaps axe manufacture holds an exclusive status
However, it could be that certain flint seams were just not amongst the flint mines. The depths miners were pre-
physically suitable for the production of the objects to be pared to dig to access particular seams and the inherent
manufactured and were thus rejected for exploitation. dangers involved may have added an implicit value to
The evidence from across Europe shows that although the axes. The questions and values associated with dif-
there was plenty of surface flint, easily and readily availa- ficulties of access, can be mirrored amongst stone axe
ble, communities within the early Neolithic were prepared production sites.
to go to great lengths to access raw material from deep Barber, 2001 has suggested regarding flint mines and
within the earth (Barber 2001: 22-23). Even the unexploit- quarry sites, that perhaps it is the location and manner of
ed seams and poorer quality seams within the mine shafts exploitation that highlights the desire to obtain stone in a
could have been utilised for the large scale production of particular way or from particular place.
routine blades, scrapers and knives, but were barely uti- The importance of the axe and its inherent association
lised. Why? Perhaps it is the association of the mine shafts with flint mines is a common thread that is shared be-
and the items produced there. tween the mines across Europe. This commonality is fur-
ther supported by the use of similar procedures in regard
4. The Choice of Product to be manufac- to the manufacture of axes at these sites. The process of
tured manufacture is always the same throughout all the sites in
the study region. The mined raw flint is chipped to form
The data from the mines within the study region indicates roughouts at the site before being transported to second-
that a commonality occurs between the objects of produc- ary sites for final polishing and finishing (Bostyn and
tion and the procedures involved in their manufacture. Lanchon 1992: 136-139; Desloges 1986; Russell 2001:
Firstly the mines of Sussex, Jablines, Bretteville and Jan- 45-48). There is very little evidence for polishing or fi-
drain are all mainly manufacturing axes (Russell 2001; nal polished products at the mines (Pull 1932: 112-114;
Bostyn and Lanchon 1992; Desloges 1986; Hubert 1974). Collet et al. 1997; Desloges 1986).
159
Yet again we find similar echoes within the chronological- Similar episodes of what appear to be deliberate depo-
ly contemporary axe factories at Great Langdale, Cumbria sitional episodes can be identified throughout the other
(Bradley and Edmonds 1993: 37-38) and Seledin, France mines of Sussex. For example, within the backfill of shaft
(Le Roux, 1971), where not only do we find similar is- 7 at Blackpatch, we find a cremation deposit associated
sues regarding difficulty of access, but the manufacturing with a flaked axe, scraper, knife and a curious charm of
procedure also follows similar patterns with the axes being worked chalk (Pull 1932: 56-58). Further down the shaft
formed to roughout on the site and transported to second- fill, lay another deposit consisting of animal bone, flint
ary polishing sites. working debris and a pile of unused flint nodules, which
The parallels in not only the products produced, but also according to the excavator were placed there (Pull 1932:
the method of manufacture across the mine sites through- 49). Within shaft 4 the disarticulated remains of an adult
out Europe, continues to imply a shared influence and ide- human femur and a jaw bone from a young person were
ology associated with the process of mining and perhaps identified. The details of these deposits appear to have
more importantly axe manufacture. been placed within the shafts, after the excavation of the
raw flint material, for a purpose.
5. The choice of deposition Purposeful deposits can be identified within the backfill
of shaft 4 at Church Hill, which consists of 15 episodic
The final cultural choice associated with flint mining, ex- events both natural and deliberate. Within these, deposits
plores how the excavated mine shafts were treated during of fine, flaked axes are identified along with various flint
and after the extraction process. Almost all the mines have working floors.
evidence of depositional activities. However, are these de- Also identified within the shaft were a series of unique
posits mere random acts of waste disposal or can we iden- geometric impressions above the gallery entrances, which
tify evidence of set predetermined behaviours amongst the cannot be paralled amongst any other mine sites (Russell
mining communities across Europe to the mines them- 2001: 87-103).
selves. At Harrow Hill, amongst shaft 21 and its various galleries
Firstly we will explore some of the evidence for deliber- similar deliberatedepositional practices can be identified.
ate depositional events within certain shafts throughout Antler picks, some in an unused condition were identified
Europe; these examples consist of certain selected deposits at the entrance to galleries (Curwen and Curwen 1926).
within shafts, which would appear to be beyond identifica- Towards the base of the shaft a broken axe was deposited,
tion as a waste disposal deposit. with the 2 pieces placed 45cm apart, with different patina-
The most documented evidence exploring the idea of at- tion (Holleyman 1937: 239-242).
titudes towards the shafts, is presented by Peter Topping Such practices associated with the backfilling of the shafts
(2005), who integrated an ethnographical framework into and deliberate deposition in the galleries can be witnessed
hypothesised formal structured depositional events with the amongst the flint mines in Western Europe.
backfill of shaft 27 at Cissbury, Sussex. The first episodes At Jandrain we find deposits of virgin unused picks and
of deposition within the shaft were associated with cleans- roughout axes within the galleries (Hubert 1974: 15). Sim-
ing rituals and offerings; subsequent deposits were associ- ilar practices are recognised at Bretteville-le-Rabet, depos-
ated with the process of extraction while later episodes of its of unused tools are also found within the shafts. Further
deposition are associated with tool production and rites of evidence from the shafts suggests tools were deposited at
renewal (Topping 2005: 78-79). The near complete skel- the base of exploited seams (Desloges 1986: 95). At Ja-
eton of an ox at 3.5m can be interpreted as an offering, as it blines, Rijckholt and Spiennes, large numbers of roughout
is an almost complete ox, while it could also be a waste de- axes have been identified deposited within shafts and gal-
posit or an unfortunate accident. Further deliberate deposits leries (Bostyn and Lanchon 1992; Felder et al. 1998; Ver-
within this shaft are the numerous axes and axe roughouts heyleweghen 1966). Further evidence of deliberate deposi-
and the deposition of the human female skeleton lying at tion can be identified when considering deposits of human
the entrance to the west gallery with what appear to be two remains. At Rijckholt firstly, a skull with no mandible was
chalk charms in her right hand (Topping 2005: 71-72). deposited at the end of one of the galleries, in a shallow pit
160
(De Grooth 1998). The gallery was later backfilled with level but within individual mine site, from shaft to shaft.
chalk, except within the area of the skull (Bosch 1979; De Individual shafts within a mine site would not only have
Grooth 1998). Very similar practices have been identified been dug by different individuals, but also differing gen-
at Spiennes. Buried skulls have been identified without the erations, kinships and perhaps cultures. Each individual
mandibles, these deposits have previously been associated could theoretically choose to represent their belief in dif-
with a specially protected phase (Verheyleweghen 1966). ferent ways and thus we find a variety of manifestations
The human remains identified in the Sussex mines how- and interpretations but all adhering to a general philosophy
ever do not echo identical practices. At Blackpatch the dis- and ideology.
articulated remains of an adult femur and human jaw bone The evidence implies that mining and quarrying were
from a young person were identified within shaft 4 (Pull largely for the construction of axes. There may have been
1932: 49), while as previously described, at Cissbury the an overall ideology that when this process was taking
full skeleton of an adult woman was found at the entrance place either certain seams of flint or the difficulty of pro-
to a gallery in shaft 27 (Russell 2001: 178-184). curement, would add prestige and a cultural value to the
The evidence indicates that throughout Europe, the min- object of production. However each culture or individual
ing communities chose to make certain deposits within the would choose at which point during the mining process
shafts, most commonly after mining and during the back- they believed they had satisfied this obligation.
fill process. Such deposits consist of either human remains The second general ideology identified across Europe is
placed in a particular way, as at Rijckholt and Spiennes, or one of giving back to the earth, where once extraction
often they consist of deposits of stone, axes or tools. How- had been completed, the individual was obliged to create
ever there does not seem to be a universal ideology, in the a deposit, an offering so to say. Again this offering could
individual deposits. One would not have expected to find a follow many different manifestations, whether in the form
regimented practice of deposition throughout Europe, and of human bone, axes themselves or tools, each individual
within each mine and individual shaft, but we can perhaps would choose how they would satisfy there obligation un-
see a general trend and attitude to the mines in which the der this general cultural ideology. Thus the range of de-
individual interpretation of how to represent a general posits can be different from site to site and shaft to shaft
belief or attitude is flexible and dynamic. We may be see- across Europe. In general though we find that both cultures
ing a general reflection of the choice and desire to give in Sussex and across Western Europe are following this
back to the earth, perhaps echoing hypotheses regarding general cultural ideology.
rights of renewal.
Conclusion
6. Discussion
The mining communities across Europe each employ a
The analysis of the cultural choices taken place within the similar set of cultural choices in their attitudes to flint min-
flint mining process supports the evidence for common ing. These common practices first show that across West-
cultural ideology in respect to the attitudes and behaviour ern Europe, cultural ideas not just in the physical mani-
towards the mining process and the mines themselves. festation of mining technology and techniques are being
Common practices have been identified throughout Eu- transferred, but perhaps the ideology and beliefs integrated
rope in regards to the desire to mine for particular flint with such activities and process are being adopted as well.
seams with implicit values and to produce similar prod- The action of extracting stone from the earth for the crea-
ucts in similar ways. This analysis has also highlighted the tion of axes appears to be integral to this common ideol-
choices made by cultures throughout Europe to make cer- ogy, which is emphasised by cultural choices and the use
tain deposits within the shafts. of structured depositional practices.
However it must be emphasised that this study is look- The discussion above outlines the possibility of a general
ing at broad generalisations and ideologies. As has been philosophy and ideology that underpins and unites Neo-
identified throughout the depositional analysis, there are lithic communities across Europe. It is stressed however
differences of interpretation not only on a micro regional that this encompasses general trends across a very large
161
area. The dynamic nature of beliefs themselves added tem- Collet, H., Deramaix, I., Sartieaux, P-P. and Vander
poral and spatial variability explains the differences in the Linden, M. 1997. Fouille prventive de puits dextraction
micro detail between regions, mine sites and shafts. de silex Petit-Spiennes (Hainaut). Notae Prehistoticae,
Ultimately however, the cultural choices identified 17, 203-212.
amongst the mines of Europe are being practiced within
the Sussex mining tradition, implying that a continental Curwen, E. and Curwen, E.C. 1926. Harrow Hill flint mine
influence may lie behind their origin, development and excavations 1924-5. Sussex Archaeological collections,
philosophy of use. 67, 103-138.
De Grooth, M.E.Th. 1998. The Flint Mines at Rijckholt-

Paul Wheeler Sint Geertruid and Their Socio-economic Interpretation.
Bournemouth University In M.Edmonds and C.Richards (eds) Understanding the
e-mail: pwheeler@bournemouth.ac.uk Neolithic of North-Western Europe. Cruithne Press. Glas-
gow, 351-369.
References Desloges, J. 1986. Fouilles de mines silex sur le site

Nolithique de Bretteville-le-Rabet (Calvados). Revue
Barber, M. 2001. Flint Mines in the early Neolithic and be- archologique de lOuest, supplement no 1, Rennes, 73-
yond. Raw material sources and later prehistory in Southern 101.
England. In A.T.Smith and A. Brooks (eds) Holy Ground:
Theoretical Issues Relating to the Landscape and Material Felder, P.J. 1981. Prehistoric flint mining at Rijckholt St
Culture of Ritual Space. Papers from a session held at the Geertruid (Netherlands) and Grimes Graves (England).
Theoretical Archaeology Group conference, Cardiff 1999. In F.H.G Engeln (ed) Proceedings of the Third Interna-
Oxford. BAR International Series 956, 21-26. tional Symposium on Flint, 24-27 mei 1979 Maastricht.
Staringia 6, 1981, 25-28.
Barber, M., Field, D. and Topping, P. 1999. The Neolithic
Flint Mines of England. Swindon. London. Royal Com- Felder, P.J., Rademakers, P. Cor M., and De Grooth,
mission on the Historical Monuments of England/English M.E.Th. 1998. Excavations of Prehistoric flint mines at
Heritage. Rijckholt-St.Geertruid (Limburg, The Netherlands) by the
Prehistoric Flint Mines Working Group of the Dutch Ge-
Bosch, P.W. 1979. A Neolithic Flint Mine. Scientific ological Society, Limburg Section. Bonn. Archaologische
American 240 (6):126-132 Berichte 12.
Bostyn, L. and Lanchon, Y. (eds) 1992. Jablines le Haut Holleyman, G. 1937. Harrow Hill Excavations, 1936. Sus-
Chteau (Seine-et-Marne): une minire de silex au No- sex Archaeological Collections, 78, 230-251.
lithique. Paris. Maison des Sciences de lhomme. 246
Hubert, F. 1974. Minires nolithiques Jandrain-Jan-
Bostyn, L. and Lanchon, Y. 1997. The neolithic flint mine drenouille en Brabant. Archaeologica Belgica, 167, 5-44.
at Jablines Le Haut-Chteau (Seine et Marne). In Sili-
ceous Rocks and Culture Ramos-Millan et M.A. Bustillo Hubert, F. 1978. Une minire nolithique silex au Camp-
Eds. Actes du VI International Flint Symposium, Madrid, -cayaux de Spiennes, Archaeologica Belgica, 210, 1-43
1-4 octobre 1991, Universit de Granada, p. 271-292.
Bradley, R. and Edmonds, M. 1993. Interpreting the axe Le Roux, C.T. 1971. A stone axe-factory in Brittany. An-
trade: Production and exchange in Neolithic Britain. Cam- tiquity, 45, 283-288.
bridge. Cambridge University Press.
162
Pull, J.H. 1932. The Flint Miners of Blackpatch. London.

Williams and Norgate.
Russell, M. 2001. Rough quarries, rocks and hills : John Pull

and the Neolithic flint mines of Sussex. Oxford. Oxbow.
Topping, P. 1997. Structured deposition, symbolism and

the English Flint mines. In R.Schild and Z.Sulogostowska
(eds), Man and Flint: Proceedings of the VIIth International
Flint Symbosium, Warszawa Ostrowiec Swietokrzyski,
September 1995. Warszawa. Institute of Archaeology and
Ethnography, Polish Academy of Sciences, 127-132.
Topping, P. 2005. Shaft 27 Revisited: An Ethnography of

Neolithic Flint Extraction, In Topping, P. and Lynott, M.
(eds) The Cultural Landscape of Neolithic Mines. Oxford.
Oxbow, 63-93.
Verheyleweghen, J. 1966. Le Nolithique minier belge:

son origine et ses relations culturelles. Palaeohistoria, 12,
529-557.
Weigsberger, G., Slotta, R. and Weiner, J. (eds) 1980. 5000

Jahre Feuersteinbergbau. Die Suche nach dem Stahl der
Steinzeit. Verofentlichungen aus dem Deutschen Bergbau
Museum Bochum 22.
Whittle, A. 1995. Gifts from the earth: symbolic dimen-

sions of the use and production of Neolithic flint and stone
axes. Archaeologia Polona, 33, 247-260.
Acknowledgments
The writer would like to thank Franois Giligny for the

kind invitation to speak at the EAA session in Cracow
and also to thank Pierre Allard, Franoise Bostyn and
Jacek Lech for organizing the session. Further thanks
to Alan Saville, Yoann Gauvry, Anne Hauzeur and
Marjorie de Grooth for all their comments and assistance
in contributing and developing ideas to the research. James
Morris also kindly read and commented upon the early
draft of the paper.
163

Flint Mining in Prehistoric Europe Inter-1 PDF

Загружено:

Сведения о документе

Оригинальное название

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

Flint Mining in Prehistoric Europe Inter-1 PDF

Загружено:

Авторское право:

Доступные форматы

European Association of Archaeologist, 12th annual meeting

Cracow, Poland, 19th-24th September 2006

BAR International Series 1891

Flint mining in the Prehistoric Europe: interpreting the archaeological records

Editing: Dominique BOSSUT

Secratary of edition: Franoise Bostyn

Flint working at the early linearbandkeramik settlement of Geleen-Janskamperveld........................... 13

A new flint mine at Flins-sur-Seine/ Aubergenville (Yvelines, France)...................................................... 79

The Krzemionki flint mines latest underground research 2001-2004....................................................... 97

Flint extraction and processing from

Keywords: Den of Boddam, flint, Neolithic, quarrying, Scotland, secondary deposits.

extractive technology, would have formed reliable sources

1. DEPOSITS AT DEN OF BODDAM

One such location where a secondary flint source was

Fig. 3: Den of Boddam location map, showing the

Samples taken from the same level of the BRG at Den of

2. QUARRYING clear signs of edge collapse (Fig. 6). This is unsurprising

Lab code Site coding and Sample type Context 13C 14

Table 1: Radiocarbon dates from Den of Boddam.

3. PROCESSING Some hammerstones have been used as anvilstones (or vice

Fig. 7: Den of Boddam. Hammerstones. Fig. 8: Den of Boddam. Anvilstones.

Archaeological investigations at Den of Boddam have re-

Saville, A., 2005. Prehistoric quarrying of a secondary flint

Saville, A., 2006. Flint technology and production asso-

Weisgerber, G., 1987. The technological relationship be-

Wickham-Jones, C.R. and Collins, G.H., 1978. The sourc-

I am grateful to Franoise Bostyn for the initial invitation

Flint working at the Early Linearbandkeramik

Marjorie E. Th. de Grooth

Type JKVE1 JKVE2 JKVE3 JKVE4 All Early

Primary Slope Residual Oligocene Oligocene

Table. 2: Geological characteristics of known extraction points.

White zone Red zone Brown streaks Dusty surface

Fig. 3: Comparison of translucency between JKV and extraction points.

Types Early habitation All material

(micro) Burins 4 0.8 5 0.6

Table. 4: JKV, the main tool types.

Table. 6: JKV, diachronic variability of main artefact types through time.

Flakes : Blades+Tools 2.6 3.4 3.4 2.3

Table. 7: JKV, proportions of main artefact categories through time.

4. Inter-site comparisons pieces with cortex, as well as a smaller amount of unre-

JKV JKV LW8 Elsloo

JKV LW 8 Early Ald. Platte, Hambach 8

The debris recovered at the Banholt extraction site makes

houses. In the next generation, not only did the number of

Zimmermann, A. 1988. Steinmaterial. In U. Boelicke, D.

Zimmermann, A. 1995. Austauschsysteme von Silexarte-

Zimmermann, A. 2002. Landschaftsarchologie I: Die

Zimmermann, A. 2006. Production and demand of flint

An economy of surplus production in the Early

Pierre Allard, Laurence Burnez-Lanotte

Keywords: Neolithic, flint, Bandkeramik, overproduction.

Experimental debitage (method 1)

The experimental debitage was carried out by Jacques Pe-

Refits (method 3) 5. Results

In Stone Age-Mining Age, VIII International Flint Sympo-

Astruc, L., Gartuze, B., Pelegrin, J. and Akkermans, P. in

Binder, D. and Gassin, B. 1988. Le dbitage laminaire

Burnez-Lanotte, L. and Allard, P. 1998. Mode de produc-

Burnez-Lanotte, L. and Allard, P. 2003. Blade debitage in

We are very grateful to the Fyssen Foundation for the in-

The prehistoric flint mining complex

Hlne Collet, Anne Hauzeur, Jacek Lech